NEW SERIES

THE CLASSICAL QUARTERLY

C
THE
CLASSICAL
QUARTERLY

VOLUME 59 NUMBER 1 2009

Cambridge Journals Online

For further information about this journal please
go to the journal web site at:

journals.cambridge.org/caq

PUBLISHED FOR THE CLASSICAL ASSOCIATION

BY CAMBRIDGE UNIVERSITY PRESS

CLASSICAL QUARTERLY
P U B L I S H E D I N M AY A N D DE C E M B E R I N E A C H Y E A R

CLASSICAL ASSOCIATION JOURNALS BOARD

Professor R.L. Hunter (Chairman)
Miss C. Davenport (Secretary)
Dr A. Bowie
Professor D. Cairns
Dr J. Davidson
Mr P. Hooker
Dr J. Morgan
Professor S.P. Oakley
Mrs C. Rouech
Professor J.H.D. Scourfield
Dr P. Stewart

Ex officio
Dr V. Izzet (Editor, Greece &Rome)
Dr R. Shorrock (Editor, Greece &Rome)
Dr J. Taylor (Editor, Greece & Rome)
Dr R. Ash (Editor, Classical Quarterly)
Professor J. Mossman (Editor, Classical Quarterly)
Professor R. Gibson (Editor, Classical Review)
Dr N. Hopkinson (Editor, Classical Review)

EDI TORS
Rhiannon Ash, Merton College, Oxford, UK
Judith Mossman, University of Nottingham, UK

The Classical Quarterly is published by Cambridge University Press on behalf of the Classical Association. It is
under the control of a Journals Board consisting of a Chair appointed by the Council of the Classical Association,
the Chair of the Council, the Secretary to the Council, the Treasurer, and six further members appointed by
Council.
CONTRIBUTIONS TO THE JOURNAL
Contributions and correspondence on its subject matter should be addressed to Dr Rhiannon Ash, Merton College,
Oxford OX1 4JD, UK, Email: cqeditor@classics.ox.ac.uk or Professor J. Mossman, Department of Classics,
University of Nottingham, University Park, Nottingham NG7 2RD, UK, Email: judith.mossman@nottingham.ac.uk.
Authors are requested to follow the Notes to Contributors printed at the back of the journal.
Subscriptions
The Classical Quarterly (ISSN 0009-8388) is published twice a year in May and December. Two parts form a
volume. The subscription price which includes delivery by air where appropriate (but excluding VAT) of volume
59, 2009, which includes print and online access, is 106.00 (US $196.00 in USA, Canada and Mexico) for
institutions. The online-only price available to institutions is 99.00 (US $182.00). Members of the Classical
Association may subscribe at reduced rates. Single parts are 57.00 (US $104.00 in USA, Canada and Mexico)
plus postage. EU subscribers (outside the UK) who are not registered for VAT should add VAT at their countrys
rate. VAT registered members should provide their VAT registration number. Japanese prices for institutions
(including ASP delivery) are available from Kinokuniya Company Ltd, P.O. Box 55, Chitose, Tokyo 156, Japan.
Orders, which must be accompanied by payment, may be sent to your subscription agent or direct to the publisher:
Cambridge University Press, The Edinburgh Building, Shaftesbury Road, Cambridge CB2 8RU; or in the USA,
Canada and Mexico: Cambridge University Press, Journals Fulfillment Department, 100 Brook Hill Drive, West
Nyack, New York 10994-2133, USA. Periodicals postage paid at New York, NY and at additional mailing offices.
COPYING
This journal is registered with the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923, USA.
Organizations in the USA who are also registered with the C.C.C. may, therefore copy material (beyond the limits
permitted by, sections 107 and 108 of U.S. Copyright law) subject to payment to the C.C.C of the per copy fee of
$15.00. This consent does not extend to multiple copying for promotional or commercial purposes. Code
0009-8388/09. ISI Tear Sheet Service, 3501 Market Street, Philadelphia, PA 19104, USA, is authorized to supply
single copies of separate articles for private use only. Organizations authorized by the Copyright Licensing Agency
may also copy, material subject to the usual conditions. For all other use, permission should be sought from
Cambridge or from the American Branch of Cambridge University Press.
The Classical Quarterly is included in the Cambridge Journals Online journals service which can be found at
http://journals.cambridge.org/caq. Tables of contents are freely available. The full text is available online from your
desktop if your library has a subscription, or if you subscribe as a member of the Classical Association. For further
information on other Press journals access http://journals.cambridge.org.
The Classical Association 2009
ISSN 0009-8388
Printed in the United Kingdom by Bell and Bain, Glasgow, UK

Classical Quarterly 59.1 17 (2009) Printed in Great Britain

doi:10.1017/S00098388090000019

1
TORRANCE
OATH AND VIRTUE ISABELLE
IN EURIPIDES
H ELEN

ON YOUR HEAD BE IT SWORN: OATH AND

VIRTUE IN EURIPIDES HELEN
It has long been recognized that Euripides Helen is a play which explores the tension
between illusion and reality in a very sophisticated and complex way.1 The real Helen
in this play is a paradigm of chastity and virtue. Not quite a Penelope, but almost.2 It
is clear that the mythic variant Euripides is dramatizing goes against tradition, and
Euripides takes great care in emphasizing Helens chastity, a quality which is antithetical to the standard character of the tragic Helen.3 Many scholars have discussed
the issue of Helens chastity in this play, especially in terms of implied parallels with
the virgin Persephone, abducted by Hades.4 Apart from being chaste, Helen is also a
clever schemer, but Menelaus, by contrast, has been treated as dim-witted, pompous
and contemptible by the vast majority of scholars.5
This paper argues that the use of oaths in the play sheds further light on the virtue
of both Helen and Menelaus. In the case of Helen, it suggests that her oath-taking
confirms her chastity and looks forward to her predicted apotheosis. In the case of
Menelaus, it suggests that the oath bond with Helen casts him in a more positive light
as a character and gives him more credit than has generally been acknowledged by
scholars.
In a drama of doubles, it comes as no surprise that there are two oaths.6 The first
comes quite early on. Helen has been devastated to hear from Teucer that Menelaus is
believed dead (132), but the Chorus persuade her to find out for sure from Theonoe,
blessed with the divine gift of knowledge, whether or not the rumours are really true
1 See e.g. C. Segal, The two worlds of Euripides Helen, TAPhA 102 (1971), 553614;
M. Wright, Euripides Escape Tragedies: A Study of Helen, Andromeda, and Iphigenia among
the Taurians (Oxford, 2005), esp. 278337.
2 The parallels between Helen and the Odyssey have been well noted. See e.g. R. Eisner,
Echoes of the Odyssey in Euripides Helen, Maia 32 (1980), 317; W.G. Arnott, Euripides
new-fangled Helen, Antichthon 24 (1990) 118, at 13; I.E. Holmberg, Euripides Helen: most
noble and most chaste, AJP 116 (1995), 1942.
3 Eur. El. 12803 is the only other instance in tragedy where Helen is presented as a blameless
victim of the gods, there explained by her brothers the Dioscuri.
4 E.g. Segal (n. 1), 595600; C. Wolff, On Euripides Helen, CP 77 (1973), 6184, esp. 634.
D.M. Juffras, Helen and other victims in Euripides Helen , Hermes 121 (1993), 4557
discusses both parallels and distinctions between Helen and Persephone.
5 Cf. recently W. Allan (ed.), Euripides: Helen (Cambridge, 2008), ad 3936 and 453; Wright
(n. 1), 198, 283, and previously A. Pippin [Burnett], Euripides Helen: a comedy of ideas, CP 55
(1960), 15163, at 158; A.M. Dale (ed.), Euripides: Helen (Oxford, 1967), xii; C.H. Whitman,
Euripides and the Full Circle of Myth (Cambridge, MA, 1974), 4550, 567, 61; Segal (n. 1), 575,
610; D.G. Papi, Victors and sufferers in Euripides Helen, AJP 108 (1987), 2740, at 39;
N. Austin, Helen of Troy and her Shameless Phantom (Ithaca and London, 1994), 141, 15762;
P. Pucci, The Helen and Euripides Comic Art, Colby Quarterly 33 (1997), 4275, esp. 5966;
P. Mureddu, Gli stracci di Menelao: polemica ed autoironia nell Elena di Euripide, Philologus
147 (2003), 191204, at 1912. Arnott (n. 2), 15 calls Menelaus a man of limited brain.
A.J. Podlecki, The basic seriousness of Euripides Helen, TAPhA 101 (1970), 40118, was
refreshingly radical in suggesting that Menelaus is a serious Homeric hero, and not a comic
buffoon, at 4023 and passim, but this view has not been popular.
6 On doubles see Segal (n. 1); G.S. Meltzer, Where is the glory of Troy?: Kleos in Euripides
Helen, CA 13 (1994) 23455; Wright (n. 1), 328.

ISABELLE TORRANCE

(30630). Helen agrees, but swears an oath, invoking the Spartan river Eurotas as
sanctifying deity, to commit suicide if the rumours of her husbands death are true
(34859). This is a particularly solemn oath statement, and Helen seriously contemplates various options of hanging and suicide by sword-stroke. The passage is a strong
marker of how we are to interpret Helen as a character. It should be absolutely clear
to the audience that Helen really does love her husband, even after all this time, and
will not consider taking another man. Whatever illusions manifest themselves in this
play, Helens devotion to her husband is not one of them. The invocation of Eurotas is
also significant. Helen chooses a deity connected to her marital home in Sparta as
overseer of her oath.7
The sanctifying deity Helen chooses for her second oath in the play, an oath of
essentially the same content, is even more striking. In fact, she does not call any deity
to witness her oath, but invokes the head of Menelaus as sanctifying power (835:
). It occurs to Helen, as she and Menelaus are
planning an escape ruse, that if Theonoe is not won over, Menelaus will be killed and
she will be forcibly married off to Theoclymenus (833). Menelaus feels that Helen is
mentioning forced marriage as an excuse (834), proving that old associations die
hard. But Helen immediately responds by volunteering this powerful oath, which, like
the first one she had taken before Menelaus arrived, clears her of any suspicions with
regard to her chaste intentions. She swears by the head of Menelaus to die and never
take a new husband should he be killed (8357). The oath is formalized with a handclasp (8389), and Menelaus reciprocates in kind stating that he will kill himself if he
loses Helen (840). They agree that in the event of disaster, they will take refuge at the
tomb of Proteus, father of Theoclymenus and Theonoe, and defend themselves for as
long as possible before committing suicide (8414).8
The formal swearing of an oath in Greek poetry usually involves the invocation of
a power greater than oneself .9 This power is normally one or more deities or, more
rarely, a sanctifying object. In cases in which a sanctifying object is sworn by, this
object generally embodies some particular power.10 A well-known example is Achilles
in Iliad 1.23346 who swears by Agamemnons sceptre. The sceptre is a symbol of
power over the army and control over the situation, something which is very precious
to Achilles at that moment, and something which he does not wish to lose. When
Helen swears an oath to her husband invoking his own head, and no other power as
sanctifying force, the implication is similar. It shows that she is desperate not to lose
Menelaus, and this is validated by her sworn statement that she is prepared to die
7 Podlecki (n. 5) suggests, at 410, that the Eurotas seems to symbolize Sparta and the happier
days there to which the principals fear they may never return.
8 Allan (n. 5) ad 835 follows M. Lloyd, The tragic aorist, CQ 49 (1999), 2445, at 312, in
arguing that Helen does not intend to swear to this death pact but that Menelaus completes her
oath for her using his own interpretation of what she should swear (at 836). However this analysis
treats too lightly Helens previous oath at 3536, which is a sworn statement to kill herself should
Menelaus be found dead, not merely a threat to do so, as Allan suggests. In fact it is not unusual
for a respondent in a stichomythic exchange to pick up and clarify the meaning of the interlocutor in his or her response since the stichomythic pattern only allows for one line to be
expressed by each person in turn; some examples taken at random are Eur. Andr. 91118, El.
5557, 6356, IT 11867, Or. 4323. Surely we are to understand that Menelaus has correctly
interpreted his wifes intentions at 836.
9 R. Janko, The Iliad: A Commentary Vol. 4, Books 1316 (Cambridge, 1992), ad 14.2719. Cf.
A.H. Sommerstein, Introduction, in A.H. Sommerstein and J. Fletcher (edd.), Horkos: The
Oath in Greek Society (Exeter, 2007), 18, at 2.
10 See also Sommerstein (n. 9), 2 and 218, n. 8 on sanctifying objects.

OATH AND VIRTUE IN EURIPIDES HELEN

should Menelaus be killed. Using the term head to signify person is not uncommon
in Greek poetry, and is particularly common in Greek tragedy, 11 but Helen 83542 is
unique among archaic and classical Greek oaths in containing the invocation of a
living mortal (or living mortals head) as sanctifying power.12 However, invocation of
the husbands head as sanctifying feature does have a divine parallel which commentators seem so far to have overlooked. This is Heras oath to Zeus at Iliad 15.3646.
There Hera swears an oath on Zeus sacred head (among other sanctifying deities) to
the effect that it was not by her will that Poseidon is harming the Trojans and helping
the Achaeans; rather, she suggests, it is his own passion which drives him on.13
The parallel between the oaths of Hera and Helen is that they both use their
husbands head as sanctifying object in an oath whose purpose is to allay that
husbands fears. The difference is that Hera is being duplicitous, while Helen is being
honest. It has been argued that Helen seems to have uttered the oath only to please
Menelaus, as mere lip-service, and that Helen would not have fulfilled her oath had,
for example, Theoclymenus appeared and killed off Menelaus.14 But this suggestion
ignores the solemnity with which an oath was treated in Greek literature and society.
There are remarkably few examples of broken oaths in Greek poetry and these
breaches are punished severely.15 In general swearers of oaths go out of their way to
avoid perjury.16
Euripides, of course, became infamous for the line given to Hippolytus in the play
of that name at 612:
, it was my tongue
that swore, but my heart is unsworn. This line was taken out of context by
Aristophanes especially (Thesm. 2756, Frogs 1012, 1471), and used to present
Euripides as a supporter of perjury. But in fact Hippolytus never breaks his oath.
11 As noted by editors; see R. Kannicht (ed.), Euripides: Helena (Heidelberg, 1969), P. Burian
(ed.), Euripides: Helen (Oxford, 2007), Allan (n. 5) ad Helen 835.
12 The dead can be invoked (as at Eur. Hipp. 307, IA 4736), and it seems that the list of sanctifying powers invoked by Demosthenes, On the Crown 208, including those who fought at
Marathon, Plataea, Salamis and Artemisium, and all who fell in all of Athens other wars, is
meant to include those who fought and survived, but these mortals are invoked in a very specific
context and not in isolation. This information has been ascertained by consulting the database of
A.H. Sommerstein, A.J. Bayliss and I.C. Torrance, The Oath in Archaic and Classical Greece
(Nottingham, 2007), accessible at www.nottingham.ac.uk/classics/oaths/database.php.
13 It has been argued that Heras oath here remains proposed but unsworn (C. Callaway,
Perjury and the unsworn oath, TAPhA 123 [1993], 1525, esp. 1718), but more convincing is the
analysis of Sommerstein, Bayliss and Torrance (n. 12) and their remarks on oath id number 409.
14 Pucci (n. 5), 64.
15 Examples include Odyssey 12.298307, where Odysseus men break their oath not to
slaughter any sheep or oxen they come upon on the island of the Sun god, and they subsequently
all die in a god-sent storm, and Hellanicus, The First Trojan War fr.26b (Jacoby), where the
Trojan Laomedon breaks his oath to Apollo and Poseidon and fails to pay them the agreed wage
for building the walls of Troy. In response Poseidon sends a sea monster which destroys those at
hand and the crops. Informal oaths can occasionally be broken without consequence, but these
belong to a different category of oaths, see A.H. Sommerstein, Cloudy swearing: when (if ever) is
an oath not an oath?, in Sommerstein and Fletcher (n. 9), 26788.
16 Several passages of Herodotus make this clear. At 3.745 Prexaspes anticipates punishment
for perjury and commits suicide. At 4.1545 Themison fulfils his oath by throwing Etearchus
daughter into the sea, but immediately hauls her out again. At 4.201 the Persians purposely swear
oaths over a hidden trench which will only be valid if the ground on which they stand remains
firm. At 6.62 Agetus is forced to give up his wife after being tricked by an oath. All these
examples were discussed by A.J. Bayliss, The artful dodging of oaths in Herodotus, at the 2007
APA meeting. The abstract is available at http://www.apaclassics.org/AnnualMeeting/07mtg/
abstracts/bayliss.pdf.

ISABELLE TORRANCE

Indeed shortly after uttering this line, which is itself spoken in a rage, Hippolytus
quickly admits that his reverence of the gods will make him refrain from breaking the
oath he took in their name (657). When Theseus returns, Hippolytus again wonders
whether he should unseal his lips in frustration at Theseus refusal to believe him, but
decides that it would have no purpose since he would at once violate his oath and fail
to convince Theseus (10603). Hippolytus piety in relation to oaths is even confirmed
by Artemis in the exodus where she praises Hippolytus not only for rejecting the
Nurses proposal, but also for keeping his oath of silence even in the face of Theseus
slanderous accusations (13069). We hear from Aristotle (Rhetoric 1416a312) that
Hygianons attempt to have Euripides charged with impiety because of Hipp. 612 was
a failure.17
There is no reason to doubt that Euripides took oaths as seriously as other
fifth-century Greeks. Instances in Euripides where oaths are perceived as having been
broken lead to disaster for the perjurer. Jason in Euripides Medea essentially suffers
the traditional punishment for perjury through the extinction of his family line (cf.
Hdt. 6.86), and Medea can be read as the personification of the avenging fury
inflicting this punishment (cf. Med. 1260).18 At Phoenissae 4812, Eteocles is
presented as a perjurer and ultimately dies, and Capaneus who swears in defiance of
the gods (Supplices 498) is killed by Zeus thunderbolt. If it were merely a question of
Helen giving lip service to Menelaus in Helen, she could easily have given a promise
rather than an oath.
Helens oath by her husbands head may recall Heras oath to Zeus in Iliad 15, but
there are several other deities who invoke the head of Zeus in oaths, and only deities
do so. Two such invocations are abortive. Hermes, in the Homeric Hymn to Hermes,
volunteers to swear an oath to Apollo on his fathers head, but never does (2746).
Indeed Hermes engagement with oaths throughout this hymn is consonant with his
tricksy character.19 Slightly different is the course of events at the end of Sophocles
Trachiniae. In the throes of death, as he supposes, Heracles asks his son Hyllus to
swear an oath on the head of Zeus his father (1185). Heracles is not yet fully divine,
but he is the son of Zeus, and there are strong arguments to suggest that we are to
anticipate that Heracles will undergo apotheosis after the end of the play.20 What is
most telling is that Hyllus, who is certainly not divine, when he agrees to swear the
oath, does not invoke the head of Zeus, he simply invokes Zeus (1188). This confirms
that swearing by Zeus head is restricted to divinities.
Two further examples of divinities invoking the head of Zeus are significant in the
context of Helen. Sappho fr. 44A and the Homeric Hymn to Aphrodite (268) both
contain virgin goddesses swearing on the head of Zeus to maintain their chastity for
17 J.D. Mikalson, Honor Thy Gods: Popular Religion in Greek Tragedy (Chapel Hill, 1991), 86,
notes the irony of Hipp. 612 being used to suggest promotion of perjury, when Hippolytus is
most loyal to oaths in the most trying and tragic circumstances. M. Dillon, By gods, tongues,
and dogs: the use of oaths in Aristophanic comedy, G&R 42 (1995), 13551, at 1434, argues
convincingly that the Aristophanic parodies of Hipp. 612 are essentially humorous rather than
loaded with accusations of impiety.
18 Pace A. Allan, Masters of manipulation: Euripides (and Medeas) use of oaths in Medea,
in Sommerstein and Fletcher (n. 9), 11324, who argues that Jason never swore an oath to
Medea, but fails to explain why Jason never denies having sworn or broken the oath as Medea
claims.
19 See further J. Fletcher, A tricksters oaths in the Homeric Hymn to Hermes, AJP 129 (2008),
1946.
20 See R. Fowler, Three places of the Trachiniae, in J. Griffin (ed.), Sophocles Revisited:
Essays Presented to Sir Hugh Lloyd-Jones (Oxford, 1999), 16175, at 16774.

OATH AND VIRTUE IN EURIPIDES HELEN

all time, Artemis in the first case, Hestia in the second. Along with Athena, these
goddesses are said to be the only three beyond Aphrodites power (h. Aphr. 735), and
it is perhaps noteworthy that Athena, the virgin goddess, was born from the head of
Zeus.21 Clearly the content of these oaths is a strong parallel to Helens oath in
Euripides play where the female swears on the head of a significant male, as the only
sanctifying power, to remain chaste. If the audience is aware that virgin deities swear
on the head of Zeus to their chastity, Helens oath becomes even more solemn.
Similarly, if swearing on the head is a feature of divine oaths, then we have a forward
glimpse in this oath towards Helens predicted apotheosis, which is mentioned at the
end of the play (16669).22 Furthermore, we are reminded that, like Heracles, she also
has a claim to be the child of Zeus, something mentioned several times in the play (e.g.
77, 81, 470, 489, 1144, 1527).
But what are the implications of this for Menelaus? There is clearly a tension
between Menelaus wretched appearance and the implication of a parallel with a
divine formula for oath-taking. There are two possible interpretations. Either we are
to find that Menelaus is made (even more) ridiculous by the parallel, or we should
find that it shows him in a worthy and virtuous light. The first interpretation would tie
in well with the opinions of critics who see the play as a light-hearted drama of the
romantic comedy type. Menelaus receives this incredibly solemn oath while being
cast as a bumbling beggar, and the effect is intended to be ridiculous. But is Menelaus
really so contemptible in the play? Critics have seen the male characters in this play as
intellectually inferior to the females. But it is often unwise to treat Euripidean
dramaturgy in terms of simple binary oppositions. An obvious example, which fails
completely in Helen, is a Greekbarbarian antithesis, where Greek is good and
barbarian is bad. Theoclymenus is a bad barbarian of sorts, but his sister is nothing
but virtue and divine knowledge. Similarly we should not allow Theoclymenus
character to colour our perception of other male characters in the play. I suggest that
Menelaus is not nearly as dim as scholars have argued, and that the solemnity of the
oath with which Helen binds herself to her husband is a cue for the audience to expect
Menelaus to act in a noble and effective way in attempting to secure their escape.
It is true that Menelaus has taken a while to come to terms with certain revelations
that the Helen he took from Troy is just a phantom and that this woman he has
come across in Egypt is actually the real Helen.23 But we can hardly find fault with
him there. Indeed, he would be far more gullible and contemptible if he had believed
everything straightaway. It is also true that Helen encourages Menelaus to abandon
his shame and flee Egypt (805), but he is not well pleased at being treated as a coward
in this way (8068). Helen rejects any idea of Menelaus killing the king. She taunts
Menelaus with the fact that killing him is impossible (809), without really ever
explaining that this is only because the omniscient Theonoe would never allow it, not
because Menelaus is a poor swordsman. In spite of her chastity, there are traces of the

21 Wolff (n. 4), 62, notes that Aphrodite is rejected in Helen by Theonoe (10067) and is
reproached by Helen (11024).
22 On Helens connection with the divine in this play, see e.g. G. Zuntz, On Euripides Helena:
theology and irony, in J.C. Kamerbeek et al., Entretiens sur lAntiquit VI: Euripide (Geneva,
1960), 20141, esp. 218, and B. Zweig, Euripides Helen and female rites of passage, in
M.W. Padilla (ed.), Rites of Passage in Ancient Greece: Literature, Religion, Society (London and
Toronto, 1999), 15880.
23 Cf. C.W. Willink, The reunion duo in Euripides Helen, CQ 39 (1989), 4569, at 501, who
comments on the shock caused to Menelaus by the circumstances of his reunion with Helen.

ISABELLE TORRANCE

familiar Helen in this character.24 She is very much concerned about Menelaus poor
appearance (554) and she is embarrassed at the idea of Menelaus begging for food
(791). But overall, Helen is presented as a virtuous alternative to the traditional
adulterous whore of tragedy. I suggest that in this alternative world of phantoms and
barbarians, Menelaus too is meant to be understood as a more worthy and virtuous
alternative to the negative portrayal he receives elsewhere in tragedy.25
In fact, his appeal to Theonoe proves more successful than Helens and he displays
considerable intelligence in improvising explanations which ease the suspicions of
Theoclymenus. Theonoe agrees to keep quiet about their escape plan because she does
, pollute the good
not want to (9991000)
reputation of [her] father, and she is concerned about pollution from her first entry
where she is involved in a purification ritual (86570). Theonoes decision can be read
as a more direct response to Menelaus threat to pollute her fathers grave through
suicide (9845), than to Helens appeals to justice and reputation. It is Menelaus who
persuades Theonoe with what touches her heart most (960), and it is he who tells
Theonoe what Helen had left out of her plea (976), that is, their sworn pact to commit
suicide over Proteus grave should their escape attempt fail. Theonoes response
proves that Menelaus arguments have been more persuasive than Helens.26 Similarly,
Menelaus shows skill in dealing with Theoclymenus. For example, when the latter
asks why the ship must be taken so far away from the shore to perform the bogus
burial rites, he promptly replies (1271)
, so
that the waves may not wash the impurities back to land, a wholly believable
explanation.
The significance of Menelaus part in the escape can easily be measured by comparing the role of Orestes in the parallel plot line of Euripides Iphigenia in Tauris.
The two plays have often been thought of as a pair because of their similarity in
structure. But there, Orestes has no input at all in the escape plan. Iphigenia is in
complete control until they reach the offstage space of the ship, where Orestes must
fend off the Taurians who are attempting to prevent them from sailing off. Menelaus,
by contrast, is not only actively involved in securing the silence of Theonoe, he also
takes over control from Helen before they leave the stage. By line 1390, he has been
washed and is dressed as a warrior. Even scholars who find Menelaus ridiculous find
him less so at the end of the play.27 Yet he is not just less ridiculous in fact he plays a
fundamental part in successfully effecting the escape. It is he who takes over from
Helen in allaying Theoclymenus suspicions (as noted above), it is he who manages to
lead the contrary bull willingly on to the ship (15678), and it is he who organizes the
ambush of the Egyptians with his men (160610), and ultimately makes good the
escape. Again, if we compare Orestes in Iphigenia in Tauris, we do not find such a
success story. Yes, Orestes fights valiantly, but it is only the intervention of Athena
that secures the safety of the Greeks who have not yet managed to leave the shore in

24

Cf. Wolff (n. 4), 77, who remarks that the new Helen assimilates the old Helen.
Menelaus in both Andromache (e.g. 3623) and Trojan Women (e.g. 10335) is presented as
effeminate and contemptible, and his character in the later play Orestes is also highly unsympathetic.
26 Cf. Wolff (n. 4), 66 and 83, who finds Menelaus rhetoric in this scene shrewd. The appeal
to Theonoe must be based on the assumption that she has not yet made up her mind on the
subject. Lines 8923 should clearly be deleted, see further Burian (n. 11) ad loc.
27 E.g. Pippin [Burnett] (n. 5), 152 and 156, feels that Menelaus regains his courage.
25

OATH AND VIRTUE IN EURIPIDES HELEN

their ship (cf. IT 13945, 141419). In Helen, the Greeks have already escaped a
significant distance before their deception is reported to the king.
In sum, Menelaus does rather well in the final part of the play, at least, and his
fortunes improve after Helen binds herself to him by oath. The formula of this oath,
reminiscent of a divine formula, confirms the virtue and potential of an alternative
Menelaus, just as it confirms the virtue and divine potential of this new Helen.
University of Notre Dame

ISABELLE TORRANCE
itorrance@nd.edu

Classical Quarterly 59.1 829 (2009) Printed in Great Britain

doi:10.1017/S00098388090000020

8
TRAGIC HONOURS AND
DEMOCRACY
PETER
WILSON

TRAGIC HONOURS AND DEMOCRACY:

NEGLECTED EVIDENCE FOR THE POLITICS OF
THE ATHENIAN DIONYSIA*
Over the course of the last decade there has been much discussion, some of it in the
pages of this journal, on the nature of the relationship, if any, between tragedy and
democracy; and in particular, on the question of whether the Athenian City Dionysia
should rightly be described as a festival of the democratic polis. The latter is a phrase
used by Simon Goldhill in his article of 1987 The Great Dionysia and civic
ideology, which rapidly became a highly influential articulation of the position that
Athens premier dramatic festival, viewed in the round as the ensemble of its framing
ceremonies and the plays performed in contest at it, reflects or indeed enacts some of
the defining preoccupations and practices of the democratic city.1 It is also a phrase
and an idea which much recent criticism has sought to undermine.2
Fortunately, the fundamental questions at stake in this debate are rich enough to
sustain not only a decade of productive disagreement on top of the twenty-odd
years of historicizing approaches to Greek drama that led up to it but also, it is clear,
further contributions.3 Given the length of the debate, measured in years and pages,
the suggestion that there are significant items of evidence not yet considered in it may
however come as a surprise. Less surprising, given that the relevant documents are
fragmentary inscriptions, a variety of evidence of which many literary students of
drama are often very wary, if not simply ignorant. It is the purpose of this paper to
introduce this material to the debate about the politics of the classical Dionysia. And
it is hoped that this may more generally encourage a better integration of the full
range of epigraphic evidence into mainstream studies of Greek drama.
* Thanks to Alastair Blanshard, Eric Csapo and an anonymous reader for extremely helfpul
comments. I gratefully acknowledge the research assistance and substantive suggestions of
Andrew Hartwig and, in particular, the generosity of Julia Shear for sharing work prior to its
publication and discussing many points of detail. Financial assistance for this project was
provided by the Australian Research Council.
1 The Great Dionysia and civic ideology, JHS 107 (1987), 5876, at 68.
2 A small selection of items from the subsequent discussion: various contributors to J. Winkler
and F. Zeitlin (edd.), Nothing to do with Dionysos? Athenian Drama in its Social Context
(Princeton, 1990), which includes a corrected version of Goldhill (n. 1); W. Connor, City
Dionysia and Athenian democracy, C&M 40 (1989), 732; C. Sourvinou-Inwood, Something to
do with Athens: tragedy and ritual, in R. Osborne and S. Hornblower (edd.), Ritual, Finance,
Politics: Athenian Democratic Accounts Presented to David Lewis (Oxford, 1994), 26990; B. Goff
(ed.), History, Tragedy, Theory: Dialogues on Athenian Drama (Austin, 1995); C. Pelling (ed.),
Greek Tragedy and the Ancient Historian (Oxford, 1997); P. Easterling (ed.), The Cambridge Companion to Greek Tragedy (Cambridge, 1997); P. Wilson, The Athenian Institution of the Khoregia:
The Chorus, the City and the Stage (Cambridge, 2000); recent criticism and response: J. Griffin,
The social function of Attic tragedy, CQ 48 (1998), 3961; S. Goldhill, Civic ideology and the
problem of difference: the politics of Aeschylean tragedy, once again, JHS 120 (2000), 3456;
R. Seaford, The social function of Attic tragedy: a response to Jasper Griffin, CQ 50 (2000),
3044; P. Rhodes, Nothing to do with democracy: Athenian drama and the polis, JHS 123
(2003), 10419; D. Carter, Was Attic tragedy democratic?, Polis 21 (2004), 125.
3 Note for instance the conference organized by David Carter at the University of Reading in
September 2007, Why Athens?: reappraising tragic politics. An edited volume of essays is forthcoming with Oxford University Press.

TRAGIC HONOURS AND DEMOCRACY

THE GREAT DIONYSIA AND CIVIC IDEOLOGY: A BLIND-SPOT

My concern is with one of the key items discussed by Goldhill namely, the proclamation in the theatre of honours, and in particular crowns, to benefactors of the city
prior to the performance of tragedy. It was Goldhill who for the first time attempted
to understand this, along with other preplay ceremonials, as part of a larger dynamic
between tragic text and context, between democratic society, politics and drama. That
discussion initiated a rich debate that continues apace. But few of these contributions
mention, and none discusses in any detail, the first and most striking example of the
phenomenon.
Goldhill found there to be a meaningful and dynamic relation between the citys
practice of awarding honours and the nature of the tragedies that followed. In other
words, the decision to timetable the event thus was in his view driven not by mere
convenience, nor even by the simple fact of the presence of a large international
audience. He saw the ensemble of ceremonial activities that took place just prior to
the performances of tragedy as powerful assertions of the norms of democratic polis
society that were then exposed to intense scrutiny in the dramas that followed:
in the interplay of norm and transgression enacted in the festival which both lauds the polis and
depicts the stresses and tensions of a polis society in conflict, the Great Dionysia seems to me an
essentially Dionysiac event.4

The practice evinced the authority and confidence of a city that was in a position to
confer significant honour and award, and it demonstrated the collective and
competitive ideology of a democracy that sought thus to encourage others to serve it
in the same way.5 By striking contrast, the scenarios of the tragedies that followed
this display so often enacted the dangers of honour-seeking, the collapse of collective
authority and the transgression of all communal limits.
The recent criticism directed against Goldhills interpretation of this ceremony
asserts, in the first place, a lack of evidence for it in the period from which tragedies
themselves survive namely, the fifth century. Thus David Carter writes: that there is
no evidence for it in the fifth [century] disqualifies it from being used by Goldhill as
evidence of a democratic context for the tragedy that we have.6 A second and more
general criticism is that there is nothing distinctively democratic about these practices,
that they are rather the sort of activity in which any polis might engage, just as tragedy
so the argument goes is a form that appealed to and addressed the polis as such,
rather than the democratic polis in particular. Well acquainted with the whole gamut
of epigraphic and literary evidence, in his case against democratic interpretations of
Athenian drama, Peter Rhodes sees the connection between such civic business as
the award of honours in the theatre and the dramas that followed as little more than
accidental.7 He does cite the key item in the dossier of evidence that I shall be
discussing the fragmentary decree on stone (IG 13 102) which records the first
known case of honours but implies merely that its date late in the fifth century and
the fact that it honours a foreigner, rather than a citizen, limit its relevance to the

Goldhill (n. 1), 76.

Goldhill (n. 1), 623.
Carter (n. 2), 89, noting that Goldhill himself seems to acknowledge the lack of good
evidence for it in the fifth century.
7 His word: Rhodes (n. 2), 112.
5
6

10

PETER WILSON

debate. He does not mention that it honoured the men who had restored the democracy by assassinating the leader of the oligarchic rgime of 411 B.C.
When IG 13 102, and a small set of its close congeners, are properly introduced to
this debate, the position alters radically. These decrees reveal the appearance of a novel
form of publication of especially important honours awarded to its benefactors by
the Athenian dmos at the end of the fifth century their announcement by means of
the heralds voice at the tragic agn of the City Dionysia. This innovation, which was
to have a long and very rich future, needs to be seen as one response to the political
trauma experienced by the democracy in the revolutions of 411 and 404 B.C. At a
minimum we can therefore say that benefactors to Athens were indeed honoured at
the Dionysia in the age of Sophocles and Euripides. But more importantly, if, as many
hold, IG 13 102 represents not only our first evidence for the practice, but evidence for
the very start of it, the circumstances of its passing throw important light on its origin
and ideological matrix. It will, moreover, emerge that this particular award of
honours to the saviours of democracy at the Dionysia of 409 B.C. stands in a tradition,
reaching back to the start of the fifth century, of proclaiming awards to tyrant-slayers
before the performance of drama at the festival (this last one pre-play ceremonial
overlooked by both Goldhill and his critics).
IG 13 102: HONOURS FOR ASSASSINS
Consideration of the significance of this first known example of honours for benefactors announced at the Dionysia has thus been entirely absent from the debate
about the Dionysia, democracy and drama. A brief discussion of this intriguing
document so familiar to students of fifth-century democracy yet never introduced
into discussion of the history of the theatre is thus in order.
IG 13 102 is a decree of the Athenian dmos, passed in the spring of 409 B.C., that
awards extensive honours to Thrasybulus of Calydon, the assassin of the oligarch
Phrynichus, architect and leading agent of the anti-democratic revolution of 411 B.C.
The document consists of three parts. The first is the substantive proposal of honours
for Thrasybulus authored by one Erasinides, a man known to have good democratic
credentials (lines 114).8 There follow two amendments, the first of which (lines
1438) grants Thrasybulus Athenian citizenship; gives him leave to seek further
benefits; confers upon him a share in a property apportionment; and awards lesser
honours to some seven or eight of his associates. The second (lines 3847) establishes
an enquiry by the Council into allegations that bribery had been used in order to
secure an earlier decree in the same connection in favour of Apollodorus of Megara.
That Apollodorus eventually had honours conferred upon him comparable to those
for Thrasybulus is clear. It is a possibility that he too was honoured with a gold crown
at the Dionysia of 409 B.C.9
Only the main decree concerns me here. Erasinides proposal is to praise
Thrasybulus of Calydon10 for being a good man towards the dmos of the Athenians

M. Osborne, Naturalization in Athens, vol. 2 (Brussels, 1982), 20.

R. Meiggs and D. Lewis, A Selection of Greek Historical Inscriptions to the End of the Fifth
Century2 (Oxford, 1969 [1988]), 263; J.L. Shear, Polis, Demos, and Revolution: Responding to
Oligarchy in Athens, 411 to 380 B.C. (forthcoming), makes the case for the award of a crown to
Apollodorus at the Dionysia.
10 Thrasybulus ethnic is not mentioned in the inscription and cannot easily be restored. It is
relatively unusual, though not unknown (cf. e.g. Agora 16, 55 with M. Walbank, Greek inscrip9

11

TRAGIC HONOURS AND DEMOCRACY

and eager to do whatever good he can (lines 68). And, in return for the good things
he has done [for the city] and the dmos of Athenians it is proposed that he be
crowned [with a gold cr]own (line 10), the value of which was very probably 1,000
drachmas.11 The next and final clause of Erasinides proposal entails the instructions
for the public announcement, at a festival agn, of the reasons for which the dmos
had awarded the crown to Thrasybulus:
IG 13 102,12 lines 1214

410/9 B.C.

stoichedon 36
h

h
1213:

h -

Velsen.

Dinsmoor
And at the agn [of the Dionysia the herald is to announce] the reasons for which [the dmos has
crowned him].

Given the state of the inscription (with 25 of the 36 letters of line 13 missing), and
given its relatively early date and thus the lack of good parallels for various of its
phrases, this part of the text of IG 13 102 should hardly be treated as definitive.13 What
can be said with certainty about this clause? In the first place, that it does indeed
specify the public announcement of honours at a festival agn. For even though the
verb ([
]) is fully restored, the reference to an agn, immediately following the
clause that stipulates the creation and award of a crown, ensures that it is a proclamation clause. There are numerous later parallels.14 Indeed, this is the first in a very
long line of such proclamation clauses in Athenian decrees (a point to which I shall
return).
h |
(lines 1112) without doubt refers to the
Similarly, the phrase h
decision to have the herald detail the reasons that motivated the award of the crown,
rather than, as is the rule in later examples, simply to state the fact of the award. In
tions from the Athenian agora, Hesperia 54 [1985], 311), to omit reference to the place of origin
of foreigners honoured in Athenian decrees. In this case the omission may have something to do
with an earlier decree or decrees in favour of Thrasybulus, though Osborne (n. 8), 18 argues
persuasively that IG 13 102 is the first or original set of honours for Thrasybulus. See also,
expressing a different view on this point, A.W. Gomme, A. Andrewes and K.J. Dover, A
Historical Commentary on Thucydides, vol. 5 (Oxford, 1981), 30910; and C. Bearzot, A
proposito del decreto ML 85 per Trasibulo uccisore di Frinico e i suoi complici, Istituto
Lombardo (Rend. Lett.) 115 (1981), 289303.
11 lines 811:
|
|
|
Most
of the restorations in this part of the decree are, in broad outline, uncontroversial.
12 Earlier editions: IG 1.59+ (Velsen and Kirchhoff ); IG 12 110 with add. p. 303+ (Hiller); Tod
86+; Meiggs and Lewis (n. 9), no. 85; cf. SEG 10 (1949), 125 (Meritt on line 13: see below); text
and important discussion in M. Osborne, Naturalization in Athens, vol. 1 (Brussels, 1981), 2830;
Osborne (n. 8), 1621; see also C. Veligianni-Terzi, Wertbegriffe in den attischen Ehrendekreten
der klassischen Zeit (Stuttgart, 1997), 312.
13 Cf. Meiggs and Lewis (n. 9), 262: The restorations are not all certain. In Wilson and
Hartwig IG I3 102 and the tradition of announcing honours at the tragic agon (forthcoming), a
case is made for a number of other possible restorations for line 13, which introduce specific
reference to the contest of tragedies.
14 See esp. A. Henry, Honours and Privileges in Athenian Decrees (Zrich and New York, 1983),
2262.

12

PETER WILSON

effect this amounts to the specification of a text for the herald to deliver or at least
something like a set of bullet-points for the same. Though its presence and function
are clear enough here, this is one feature of this early honorific decree that does not
catch on.15 The fact that it does appear here demonstrates the importance to the
framers mind of having the reasons for the honours for Thrasybulus fully and very
publicly proclaimed, and before a larger and more diverse audience than that of the
Assembly.
It has always been assumed, no doubt correctly, that the clause also envisages the
actual act of crowning Thrasybulus in the theatre, and not simply the announcement
of the reasons for this crowning. (In this, as in many other aspects of this decree, we
might have wished for a more competent secretary than Lobon.)16 The event is thus to
be no mere report, but a live performance, complete with a script for a herald to
deliver that voices the will of the dmos.
Nor can it be doubted that the agn in question (line 13) was part of a major
Athenian polis festival. In fact, despite the loss of much of line 13, there is virtually no
doubt that the festival in question was the City Dionysia of 410/9 B.C.17 Only
one other candidate has ever been aired the athletic agn of the Panathenaea of
409/8 B.C. This was proposed by Dinsmoor in 1931, with the following text:18
h
h

h -

While this does have the advantage of specifying the relevant agn more clearly than
other restorations, little more can be said in its favour. The fact that 409/8 B.C. was
a year for the Small Panathenaea speaks very strongly against it.19 In any case,
15 R. Osborne, Inscribing performance, in S. Goldhill and R. Osborne (edd.), Performance
Culture and Athenian Democracy (Cambridge, 1999), 34158, at 3545. It does however also
appear in the next datable example of an honorific decree with proclamation clause, IG 13 125,
lines 239 of 405/4 B.C., on which see further below. And cf. SEG 29 (1979), 86 (= IG 22 20+), lines
1417 (of 393 B.C.) with D. Lewis and R. Stroud, Athens honours King Euagoras of Salamis,
Hesperia 48 (1979), 18990: Euagoras of Salamis apparently praised and honoured as (among
other things) a Hellene serving Hellas. The herald is to proclaim as much at the tragic agn of the
City Dionysia.
16 Even in the majority of cases, where decrees do not, as here, specify the enumeration of
reasons for the award, but simply require the herald
to announce the
crown the meaning is rightly understood to be to announce the award of the crown as it is
being conferred (assuming the honorand is present). See below on the honorific decree for
Epicerdes of Cyrene, the provisions of which for the announcement of an earlier crown in
addition to the current one indicate a normative association of announcing and actual crowning.
This cannot have been a hard-and-fast rule, however, as the case of Callias of Sphettus shows
known not to have been in Athens when honoured: T.L. Shear, Kallias of Sphettos and the Revolt
of Athens in 286 B.C. Hesperia Supplement XVII (Princeton, 1978). Similarly, some of the many
foreign honorands in later decrees are likely not to have been able to come to Athens for the
scheduled ceremony.
17 This is now the consensus omnium: see e.g. Meiggs and Lewis (n. 9), 263; Henry (n. 14), 30;
J.L. Shear, Polis and Panathenaia: The History and Development of Athenas Festival (Diss.,
University of Pennsylvania, 2001), 414.
18 W. Dinsmoor, The Archons of Athens (Cambridge, MA, 1931), 346, n. 6; SEG 10 (1949), 125.
19 Although the restoration of democracy took place several months before the Great
Panathenaea of 410 B.C., making that festival potentially available for this purpose, the framers of
this plan to honour Thrasybulus and his associates did not bring their action forward in time to
make use of that festival to proclaim the honours, probably at least in part as a matter of deliberate choice rather than through simple force of circumstance. See below p. 15.

TRAGIC HONOURS AND DEMOCRACY

13

Dinsmoors suggestion of Athenas festival in preference to Dionysus was motivated

by a misunderstanding of the relevant calendrical circumstances. In that respect,
Meritts calculations, which take full account of the confusion that affected the
prytanic year in 411 B.C., are to be preferred. These put the start of the year 410/9 B.C.
on 26th Thargelion 411 B.C., thus leaving ample time for the eighth prytany to propose
the decree, since the Dionysia will have begun more than twenty days into their
watch.20 In her full study of the documentary evidence for the Panathenaea over all
its recorded history, Julia Shear rejects the association of IG 13 102 with the Panathenaea.21
Dinsmoor also proceeded in the belief that the selection of festival in such proclamation clauses was motivated by little more than accidental temporal proximity: that
the next major polis festival to come after the decision had been passed by the
Assembly (whether Dionysia or Panathenaea) was the festival for the announcement
of the honours, simply because it was the next.
This belief shared, more or less explicitly, by many scholars22 deserves some
further consideration and critique. For it proceeds on the tacit and inherently weak
assumption that the framers of such important proposals were fundamentally
constrained by circumstance rather than at liberty to choose time and place. While
there can be no doubt that pressure of events and questions of proximity did play
some role in the timetabling of such announcements, in the case of IG 13 102, liberty
rather than constraint is demonstrable. It was not by accident that the Dionysia was
chosen in preference to the Panathenaea, or any other major civic occasion, as the
moment at which to crown the killers of Phrynichus.
We should give more credit to the forethought of the framers of these proposals
and look at the question from (as it were) the beginning, rather than the end, of the
process. For it makes more sense to suggest that the timetabling of proposed honours
as agenda items for the Council and Assembly was significantly influenced by the

20 B. Meritt, The Athenian Calendar in the Fifth Century (Harvard, 1928), 98 and esp. Meritt,
Athenian Financial Documents of the Fifth Century (University of Michigan, 1932), 1046.
Dinsmoor calculated that the start of the prytany of Hippothontis, during which IG 13 102 was
passed the eighth of the year was around Elaphebolion 13, the day on which the Dionysia
should have ended. The place of the Hippothontid prytany is established by Meiggs and Lewis
(n. 9), no. 84, line 27 (= IG 13 375).
21 Shear (n. 17), 414; cf. also Meritt 1932 (n. 20), 105, n. 3. The proclamation of honours at the
Panathenaea is only securely attested over half a century later (in 347/6 B.C.: IG 22 212, lines
2044; P. Rhodes and R. Osborne (edd.), Greek Historical Inscriptions 404323 BC (Oxford,
2003), no. 64), and the Dionysia far eclipses the Panathenaea (its nearest rival) as the premier site
for the announcement of such awards over the course of the entire classical period, and beyond.
The honours awarded at the Panathenaea in 347/6 B.C. are for Spartocus, Paerisades, Apollonius,
sons of Leucon ruler of Bosporus. This decree shows the further innovation of the repeated
award of crowns at successive Great Panathenaea (lines 246 with Rhodes and Osborne [n. 21],
323), in full knowledge that the crowns will never leave Athens (lines 336). The Dionysia and
Panathenaea are the two festivals at which such awards are announced with increasing frequency
over the course of the following decades. The two festivals maintained this status for some two
centuries, until around 220 B.C., at which time the Eleusinia and Ptolemaea joined them to form a
fairly stable group: Shear (n. 17), esp. 41819. Agora 16, 225, lines 911 (of 224221 B.C.) is the
first appearance of the quatrain of festivals. Shear (n. 17), 41323 traces the history of the developments in the award of crowns at the Panathenaea in full, down to the first century.
22 Cf. Shear (n. 17), 414: the timing of the decree may have determined the festival at which
the crown was announced: the Great Panathenaea of 4109 had already been celebrated at the
beginning of the year, while the City Dionysia had not.

14

PETER WILSON

anticipated advent of major festivals, and to acknowledge that a large degree of

choice was often open to their proposers as to when they initiated their action.
Concern for speedy proclamation of honours at a forthcoming festival gathering
becomes quite unambiguous when the decrees themselves explicitly stipulate that the
relevant occasion is to be the next or the imminent event. But even such instructions
for expeditious timetabling are not evidence of constraint tying the hands of their
proposers. And in fact these two similar qualifying phrases are somewhat different in
is especially
function. The use of the imminent agn
striking. It appears in another early close congener to IG 13 102, and deserves brief
mention here in its own right. This is the second instance surviving from the fifth
century of awards proclaimed at a city festival, and follows the awards for Thrasybulus by just four years. Such public proclamations of honours were doubtless still a
novelty, a high and distinct honour. This is the decree awarding honours to Epicerdes
from Cyrene, dated to 405/4 B.C. (IG 13 125).23 Epicerdes had some time earlier given
the Athenians 100 minas in their hour of need to rescue Athenian soldiers from
starvation in Sicily, and been granted a crown for it, though as the document makes
clear, he had not actually been publicly awarded this crown. He subsequently made a
further cash gift of a talent, for which, among other honours, another crown was
forthcoming. The surviving decree stipulates that [the herald] should make the
announcement, with the additional proclamation at the imm[inent agn] in the city
|
|
(
) [that earlier Epic]erdes of Cyrene [gav]e the Athenians [100 minas] for the
rescue in return for which they [crowned him for his bravery] and good-will [to the
Athenians]: lines 239.24
Like IG 13 102, this decree was probably passed some time in the eighth prytany,
during which the City Dionysia fell.25 Though precision is impossible, the imminent
agn in the city may thus have been no more than days away. The honours publicly
awarded to the Cyrenean benefactor and saviour of Athenian manpower were thus
certainly publicized at the citys greatest theatrical festival. And the remains of the
decree directing this show traces of the urgency with which this deadline was met. The
clearly works as an URGENT BUSINESS label.26 But the fact
phrase
23 Bibliography on this text, following the discovery and incorporation of a new fragment in
the excavations of the Agora in 1970: B. Meritt, Ransom of the Athenians by Epikerdes,
Hesperia 39 (1970), 11114; Henry (n. 14), 301; W. West, The decrees of Demosthenes Against
Leptines, ZPE 107 (1995), 23747, esp. 2427; D.M. MacDowell, Epikerdes of Kyrene and the
Athenian privilege of Ateleia, ZPE 150 (2004), 12733. See also A. Bielman, Retour la libert.
Libration et sauvetage des prisonniers en Grce ancienne. Recueil dinscriptions honorant des
sauveteurs et analyse critique. tudes pigraphiques, vol. 1 (Lausanne, 1994), 37 for text and commentary.
24 This decree thus, unusually, directs the proclamation of a crown that had already been
granted some time earlier (Meritt [n. 23], 11314) perhaps because Epicerdes had been unable
to receive it in person at the earlier date, though it remains a possibility that he had been awarded
the earlier crown but not the act of its public proclamation. The (lost) decree sanctioning the
earlier crown should date to around 413411 B.C. and would thus antedate the award for
Thrasybulus by a few years. Julia Shear points out to me that the abolition of democracy late in
412 may have interrupted the award of this first crown. But whether the earlier decree offered
proclamation of the crown or not, it is clear that the earlier crown was not in fact proclaimed
until 404 B.C.
25 See the calculations of Meritt (n. 23), 114, arguing for Erechtheis as the tribe in prytany.
26 The phrase appears in various Attic inscriptions in the context of e.g. the urgent appointments of ambassadors, special commissioners and the like; important announcements by
heralds; urgent elections and decisions: cf. e.g. IG 13 21, line 5 of 450/49 B.C.; IG 13 40, line 47 of

TRAGIC HONOURS AND DEMOCRACY

15

that such special instructions were stipulated to ensure that the honours for Epicerdes
were proclaimed at the imminent Dionysia does not mean that this festival was chosen
for nothing more than its proximity. On the contrary, the reverse is more plausible:
namely that the proposal was shepherded through the Assembly and carefully
formulated in this way with a view to the advent of the Dionysia. The desire for
publicity at the fast-approaching festival is clearly paramount, and has created the
impression of a last-minute rush in the business of the Assembly. Perhaps this usage,
unique in honorific decrees, reflects a real situation not far from the imaginary one,
projected on to the fifth century by Lucian, where the seven gold crowns that it is
proposed be awarded to Timon of Athens are to be proclaimed at the Dionysia, at
the time of the new tragedies, today. (Timon 51).
to identify a
IG 13 125 provides the only example of the use of
forthcoming festival in the proclamation formula of an honorific decree. It is
somewhat different from the practice, attested only from the late fourth century and
then, not in the city of Athens of specifying that such proclamations are to take
place at the first (= next) Dionysia.27 As Angelos Chaniotis has recently observed of
this phenomenon, [i]t was not self-evident that the honours were to be announced on
the next occasion.28 While evincing a concern for speedy proclamation, this usage
thus also shows that there was no inherent assumption that the next urban agn,
whatever that might be, was the moment at which such proclamations would automatically be timetabled.29
The choice of festival for the announcement of honours, and of an event within a
festival programme, was thus not neutral. And it certainly was not so when, as in IG 13
102, the practice was in the very earliest stages of its formation. Moreover, in her fine
discussion of the decree of Demophantus and its associated oath, Julia Shear has
446/5 B.C., or indeed in IG 13 102 itself, line 23, also demonstrating its ready availability at this
early period; IG 22 16, line 10 of 394/3 B.C.; IG 22 112, lines 67; IG 22 114, lines 67. However in
the great majority of uses it appears as an adverbial phrase with verbs such as
,
,
,
and not, as in IG 13 125, attributively between the repeated article
and
; cf. L. Threatte, The Grammar of Attic Inscriptions, II: Morphology (Berlin/New
York, 1996), 409.
27 For a lively discussion of this phenomenon and further examples see now A. Chaniotis,
Theatre rituals, in P. Wilson (ed.), The Greek Theatre and Festivals: Documentary Studies
(Oxford, 2007), 4866, at 567. Others: Priene, IPriene 17, line 17; 54, line 65; 53I, line 34; 53II,
line 59; 60, line 17; 61, lines 367 (fourth to second centuries); Cos, SEG 35 (1985), 912, line 5
(second century); ED 133, line 34 (second century); Larisa, SEG 26 (19767), 677, line 80 (second
century); Magnesia, IMagn. 102, line 12; Lampsacus, IK 6, 2, 12, line 11 (second century). More
elaborate variants: Colophon, REG 1999: 2, lines 2932, (third century); Ephesus, IEph. 1390,
lines 56. A (rare) Panathenaic example: IMylasa 632, lines 1920 (secondfirst centuries).
28 Chaniotis (n. 27), 56.
29 The only (late) classical Attic example known to me is from a deme, the decree of the
Aixoneis (SEG 36 [1986], 186, of 313/12 B.C.) which specifies as the time for the award of gold
crowns to two local chorgoi in the theatre at the comedies in the year after Theophrastus
archonship (lines 67:
|
). This
example further demonstrates the weakness of mere temporal proximity as the motivating factor
for the timing of proclamations. The local Dionysia of Aixone was doubtless chosen by the
Aixoneis for reasons of local pride, and because it was the one deme event that drew an audience
from well beyond the deme itself. The specification of the comedies (cf. SEG 36 [1986], 185, lines
1415 also of 313/12 B.C.) seems to reflect a special local tradition, or the performance with which
these honorands had been associated as chorgoi (or both). See D. Whitehead, The Demes of
Attica (Princeton, 1986), 23542. A second Attic example is the much later IG 22 1227, lines 313
(of 131/0 B.C.) from Salamis: a gold crown to be awarded at the tragedies of the Dionysia on
Salamis, when it next takes place (
).

16

PETER WILSON

noted that that decree was passed sufficiently early in the civic year 410/9 to have
made the Great Panathenaea of 410 B.C. available for the act of collective oath-taking
by the Athenians which it prescribed, but the Dionysia of 409 B.C. the very festival
at which Thrasybulus was crowned was explicitly specified instead.30 There could be
no more pertinent example of the deliberate selection of one polis festival in
preference to another; no better indication that simple proximity is an insufficient
explanation of such timetabling; and no more compelling evidence that the collocation of such civic rituals within a festival frame can hardly be deemed accidental.
AT THE TRAGIC AGN
It is clear that careful thought went into the plan to crown Thrasybulus at the City
Dionysia of 409 B.C. But if we return to examine the phraseology of the relevant
clause, the instruction that the honours for Thrasybulus be announnced at the agn
[of the Dionysia] might seem to be rather unhelpfully vague. The programme of
events of the City Dionysia in the late fifth century was after all complex, and the
simple expression at the agn of the Dionysia is decidedly imprecise for the practical
purpose of identifying a particular moment within it, with the aim of timetabling an
important announcement. In 409 B.C. there were at least four agnes at the City
Dionysia: the boys chorus, mens chorus, comedy and tragedy. It is, none the less,
clear that by this expression those who discussed and drafted IG 13 102 in fact meant
to be understood, at the tragic agn of the Dionysia. The close association of this
new practice from the outset with the performance of tragedy cannot be doubted.
That this is so is shown in the first place by the fact that, in the overwhelming
majority of references to the practice, literary and epigraphic, the additional specification of tragic is in fact present.31 Moreover, the imprecise and the specific
alternatives are put to the same purpose in exactly parallel epigraphic contexts.32 And
it is further suggested by the generally high profile attaching to tragedy among the
performance-genres of the festival, such as to set up a synecdochic relationship
between the agn and the tragic agn of the Dionysia. Moreover in 409 B.C., and for
many years before that date, a range of announcements and other activities the
preplay ceremonials illuminated by Goldhill customarily took place just prior to
the start of the tragic contest of the City Dionysia, marking that point above all
others as the most significant focus of collective, para-theatrical attention within the
festival programme. An abbreviated reference of the sort found in IG 13 102 the
agn of the Dionysia must depend on a widespread familiarity with these practices
that would remove any ambiguity. The display of imperial tribute in the orchstra
30 J.L. Shear, The oath of Demophantos and the politics of Athenian identity, in A.
Sommerstein and J. Fletcher (edd.), Horkos: The Oath in Greek Society (Exeter, 2007), 156. See
further below p. 26 on the significant coincidence at this festival of honours for the assassin and
oath-taking against anti-democrats.
31 Wilson and Hartwig (n. 13) assembles the evidence and makes a case for a number of other
possible restorations for line 13 of IG 13 102 itself, which introduce specific reference to the
contest of tragedies.
32 As an example of the imprecise usage of agn but with certain reference to the Dionysia
(see above p. 14) cf. the (securely) restored phraseology used in the decree of honours for
Epicerdes, IG 13 125, lines 239 (405/4 B.C.):
|
|
. Of the numerous specific usages, cf. the honours for
Asclepiades of Byzantium, to be awarded at the agn of tragedies of the Great Dionysia:
|
(IG 22 555, lines
67, of 307/6 304/3 B.C.).

TRAGIC HONOURS AND DEMOCRACY

17

prior to the performance of tragedy very probably took place in 409 B.C., as it had
done in the 420s and probably for some decades before that. In fact, Julia Shear has
recently suggested that 409 B.C. was the very year in which the display of tribute was
resumed after its suspension in 413 B.C.33 Similarly, the other activities known to have
taken place before the performance of tragedy in the fifth century are very likely to
have been present in 409 B.C.: the parade, in armour provided by the city, of young
men orphaned by war and reared by the state;34 the pouring of libations by the ten
stratgoi;35 and finally, the possible proclamation of rewards to those who murder
would-be tyrants.36 I shall return to this last item not included in most recent
discussions shortly. For it is, I suggest, an immediate antecedent to the oath of
Demophantus against anti-democrats and tyrants (assimilated clearly in that document for the first time), an oath sworn at this same festival. And the benefits the oath
of Demophantus promised to political assassins were instantiated by the proclamation of honours for Thrasybulus at the same event. It should at any rate be clear
that, in the epigrammatic language of the decree, at the agn of the Dionysia serves
as shorthand for at the tragic agn of the Dionysia.
DEMOCRATIC HISTORY AND THEATRE HISTORY
Discussion of IG 13 102 has traditionally tended to focus on the major issues of
historical significance that lie behind it. Of prime concern has been its relation to the
seismic events that led to the downfall of the oligarchic rule of the Four Hundred in
411 B.C., and to the other sources that report them. And the evidence it provides for
the nature of engktsis (the right of a non-citizen to own real property in Attica) in
the early period has loomed large in the traditions of Athenian legal and diplomatic
history;37 as have its provisions more generally for the award of honours, for this is the
first known instance of the award of a gold crown to civic benefactors in any
context.38
Analysis of the provisions for and significance of the announcement of Thrasybulus crown has been limited. While their place at the very head of the long Athenian
33 Shear (n. 30), 156. The display in the theatre is likely to date from as early as the transference
of the League treasury from Delos to the Acropolis in 453 B.C., and generally to be coincident
with the period of empire during which tribute-paying states were required to bring their
payment to Athens at the time of the Dionysia. The best evidence associates it with the 430420s:
Ar. Ach. 496508 and ad 504; Isoc. On the Peace 82; [Xen.] Ath. Pol. 3.2; see S. Goldhill, The
Great Dionysia and civic ideology, JHS 107 (1987), 5876, at 602. That it took place just prior
to the tragic agn the source which gives most precision in the issue (Isocrates) writes only in
the orchstra, at the Dionysia, when the theatre was full is guaranteed by its close association
with the parade of war-orphans, whose reception is placed precisely at the time of the tragic
performances in the city by Aeschines, In Ctes. 154.
34 This too is associated with the period of empire, and seen as long in the past in 330 B.C.:
Aeschin. In Ctes. 1534; Thuc. 2.46.1; Lys. fr. 129 (Carey): note the detail here of announcement
by a herald,
; perhaps Solonic: Diog. Laert. 1.55; cf. Rhodes (n. 2), 111.
35 Plut. Cim. 8.79 with Goldhill (n. 1), 60; Carter (n. 2), 6.
36 See below, p. 26 for the evidence for this last activity. It may also be relevant that, according
to most reconstructions of the programme of the classical Dionysia, tragedy could be said to be
the last and climactic agn, at least the last to begin (comedy perhaps having for a period also
ended on the last day): E. Csapo and W. Slater, The Context of Ancient Drama (Michigan, 1995),
107.
37 Meiggs and Lewis (n. 9), 2623 with further bibliography and Osborne (n. 8), 1621 offer
the best summary of the main issues that have occupied commentators.
38 Henry (n. 14), 30; Osborne (n. 15), 3545; Shear (n. 17), 414.

18

PETER WILSON

tradition of proclaiming honours in a live context beyond the Assembly has been
noted, it was only in 1999 that Robin Osborne associated this apparent innovation
with the particular individuals being honoured political assassins and with the
circumstances of their honouring: Announcing the honouring of Thrasybulus of
Calydon at the Dionysia is one of the distinctly novel, or at least new-fangled, ways in
which the murderers of the oligarch Phrynikhos were praised.39 He did not develop
the point further, however, and the politics of the theatre implied by it remain to be
explored. In a fine recent study, Julia Shear has done a great deal to highlight the
coincidence at this Dionysia of a number of enormously significant political events:
the award of honours for Phrynichus assassins; the taking of the oath of Demophantus; and the likely resumption of the display of tribute, after a break of some
four years in the wake of the Sicilian disaster. All of this no doubt made of that
festival a particularly charged affair.40
For this famous document, of major importance to the history of Athenian
democracy, has never been properly introduced into discussion of the history of the
Athenian theatre. It certainly deserves a place there. The reasons for that exclusion are
not hard to find. Until recently, the paths of the political history of Athens and the
history of its theatre rarely met and certainly not in the pages of PickardCambridge, who paid scant attention to the proclamation of public honours prior to
the performance of tragedy at the City Dionysia.41 For that reason alone IG 13 102
never became an item in the standard portfolio of evidence for the history of the
classical theatre that The Dramatic Festivals of Athens claimed to be. That omission,
and the very authority of Pickard-Cambridge, account for its having remained
outside the mainstream of scholarship on the theatre, even now that studies attentive
to the social and political contexts of Athenian drama have proliferated.42
Having taken a closer look at the relevant document, we can return to the broader
question of its significance for our understanding of the politics of the festival with
which I began. As Robin Osborne saw, it is clear that this new form of festival proclamation of honours for the assassin of the oligarch was an innovation tailored to the
importance of the events, giving the whole practice a profoundly democratic
origin.43 More can be said about this fitting of the new medium to the message of the
moment. For, just as the great attention paid to the public and highly visible
39 Osborne (n. 15), 354. See also M. Burzachechi, Doni ospitali (xenia) e corone doro nei
decreti della Grecia antica, Rendiconti della Accademia di archeologia, lettere e belle arti 36
(1961), 103113, at 108; M. Blech, Studien zum Kranz bei den Griechen (Berlin/New York, 1982),
1567.
40
Shear (n. 30), 156. Shear argues, however, that the oath associated with the decree of
Demophantus was taken not in the theatre but in the Agora.
41 His only reference to the epigraphic evidence is the comment at A. Pickard-Cambridge, The
Dramatic Festivals of Athens2 (Oxford, 1988 [1968]), 59, n. 3 (cf. 67): This provision occurs in
many inscriptions. H. Mette, Urkunden dramatischer Auffhrungen in Griechenland (Berlin,
1977), 94102, improves upon Pickard-Cambridge by collecting evidence for the proclamation of
honours at the Dionysia, but he too fails to include the earliest examples discussed here.
42 A similar neglect has attended the intriguing set of inscribed decrees of the Assembly,
passed at the special meeting held
after the festival, and honouring foreigners for
their services to the Athenian theatre. S. Lambert, Polis and theatre in Lykourgan Athens: the
honorific decrees, in A. Matthaiou and I. Polinskaya (edd.), :
Michael H. Jameson (Athens 2008), 5385 redresses this neglect
admirably. Cf. more generally P. Wilson, Introduction: from the ground up, in Wilson (ed.)
(n. 27), 117.
43 Osborne (n. 15), 354, quoted above. Cf. Burzachechi (n. 39), 109.

TRAGIC HONOURS AND DEMOCRACY

19

(re-)inscription of laws in the period following the turmoil of the two oligarchic
revolutions was evidently designed to bolster the authority of the democratic lawcode itself,44 so too the new and special form of non-graphic publication at the
Dionysia of honours that had been decreed by the dmos should be seen as a way of
endowing these decisions with a heightened authority that likewise takes its origins
from the (first) oligarchic trauma an authority derived moreover from a new combination of the spoken and the written words of the dmos. For the announcement in
the theatre freed many from the need to read the decree itself.45
IG 13 102 could hardly present a more redolent example, a more momentous set of
circumstances: the action that led to the restoration of democracy after its first
overthrow, a re-foundational moment in democratic history indeed, the point at
which, in the view of some, an avowedly democratic political self-consciousness was
born for the first time in Athens.46 This should be seen to confirm in spectacular
fashion Goldhills thesis of the democratic ideological frame of tragedy, for here at
the very inception of the practice, we see the democratic city rewarding with significant material gifts and powerfully symbolic honour those who came to its defence,
latter-day tyrant-slayers akin to those founding heroes of the fifth-century democracy, Harmodius and Aristogeiton.47
The international character of the Dionysia ensured that the message reached a
very wide audience. The wider world was called upon to witness the democracy once
again fully in charge of its own destiny, and, more specifically, letting all know that
anyone who did the dmos good, be they Athenian or, as with Thrasybulus, nonAthenian, would find glory and generous reward.48 The Athenian (democratic) dmos
44 M. Ostwald, From Popular Sovereignty to the Sovereignty of Law: Law, Society and Politics
in Fifth-Century Athens (Berkeley/London 1986), 50924; P. Liddel, Civic Obligation and
Individual Liberty in Ancient Athens (Oxford, 2007), 80.
45 I note that Liddel (n. 44), 172, makes a similar suggestion, to the effect that announcement
of honours made them accessible to those who did not read inscriptions. J.L. Shear, Cultural
change, space, and the politics of commemoration in Athens, in R. Osborne (ed.), Debating the
Athenian Cultural Revolution: Art, Literature, Philosophy and Politics 430380 B.C. (Cambridge,
2007), 91115, at 114 writes of Athens in 410 B.C. that democracy and its products had to be
displayed and the city had to be made visibly democratic again. She amply demonstrates how, in
preference to the Acropolis, the Agora was newly turned to this purpose, in particular for the
honouring of citizen euergetai. Between Acropolis and Agora, the theatre of Dionysus was in this
period evidently the place of choice for honouring non-citizen euergetai of democratic Athens.
46 Shear (n. 45), 114.
47 Cf. the prominent place of the Tyrannicides in the fifth-century popular imagination as
evidenced by the Harmodius skolia in particular, PMG 8936.
48 The debate around the announcement ceremony at the Dionysia has not always been clear
as to the relevance of the distinction between awards for citizens and non-citizens. In an updated
version of his influential 1987 contribution (n. 1), Goldhill (in Winkler and Zeitlin [n. 2]), 105, n.
26 very properly clarifies that the evidence for the presentation of the crowns to Athenian
citizens in the fifth century is much less secure than for the fourth century. The evidence
shows a complex picture, with awards to non-Athenians very much more prominent in the
epigraphic record overall (see esp. Henry [n. 14]) and the question of whether Athenian citizens
could legally have honours proclaimed anywhere outside the Assembly (such as at the Dionysia)
becoming a key issue at stake in the legal battle between Demosthenes and Aeschines in the 330s
B.C. (Aeschin. 3.32, Dem. 18.1201). On the last point E. Harris, Law and oratory, in
I. Worthington (ed.), Persuasion: Greek Rhetoric in Action (London, 1994), 13050, has convincingly shown that Aeschines case against the crowning of Demosthenes was, in strictly legal
terms, weak and tendentious. For the award of crowns to Athenians in the fifth century see
M. Gygax, Plutarch on Alcibiades return to Athens, Mnemosyne 59 (2006), 481500. The very
limited evidence for this includes nothing to suggest that any was proclaimed at a festival. It is
relevant to the case of Thrasybulus that in the first rider of IG 13 102 (lines 1516), he is awarded
citizenship.

20

PETER WILSON

had regained its all-important power to thank and to confer honour. And there will
also have been an important internal dynamic to the message that this proclamation
sent a clear statement to Athenian citizens, of whatever ideological persuasion, of
what constituted doing the dmos good (more on this below).
At this point it is worth passing in review the small cluster of early congeners to IG
13 102, decrees in which the festival publication formula makes its first appearance
all likewise absent from previous discussions of the politics of the theatre. For the
honours for Thrasybulus are in fact the first in a small group dating to the last decade
of the fifth and the first of the fourth century. The clear implication is that this was
one response of the democracy to the trauma inflicted by the revolutions of the Four
Hundred and Thirty a response which took root as a habit, spread far afield, and
persisted for centuries. But as I have already suggested, while the new form of
proclaiming crowns and other honours at the Dionysia was indeed a likely innovation
of this tumultuous period, seen in combination with the oath of Demophantus, the
practice should also be understood as a form of ritual renovation, a continuity with a
pre-existing practice that may date to the very first years of the democratic Dionysia,
as well as an important development upon it under pressure of very particular historical and political circumstances.
The language of these early decrees, particularly that of the proclamation clauses,
reveals the uncertainty of a formula in formation, even granted the extremely fragmentary nature of the relevant inscriptions.49 Announcement at the agn of the
Dionysia in 409 B.C is followed at a period of four years by at the imminent agn in
the city. It is clear that this has not become in any sense a fixed phrase since only two
years later we find the use of the temporal clause at the Dionysia, when the tragic
agn takes place. A new variant on this last resurfaces in the next surviving example a
decade later at the agn when the tragoidoi are in competition.50

49 See Henry (n. 14), 303. For convenience, I list the relevant clauses as restored in the most
recent editions:

IG 13 102, lines 1214 (409 B.C.):

IG 13 125, lines 239 (405 B.C.):
|
IG 22 2 frag. b, lines 1012 (403 B.C.):

|
|
|

SEG 29 (1979), 86 (= IG 22 20+), lines 1416, 2930 (393 B.C.), slightly modified:
|
36
(lines 1416); - - - - - - |
|
.
This improved reconstruction of the text of SEG 29 (1979), 86, lines 2931 (of 393/2 B.C.) follows
the recognition of Lewis and Stroud (n. 15), 189 that we have in this document a decree followed
immediately by an amendment made in the Assembly, and that the language and phraseology of
the text at lines 29ff. (amendment) closely echo that of lines 14ff. (proposal). Hence the
introduction in line 30 of the verb
from line 15, in preference to the anaemic
of IG 22 20, line 7. Lewis and Stroud (n. 15), 190 compare IG 11. 4 664,
lines 1113 (Delos, 240230 B.C.)
|
, as the nearest
parallel for the use of the phrase
. IG 11. 4 1043, lines 1516
(also from Delos) may be superior, since it is somewhat earlier and refers specifically to tragedy.
50 See the preceding note on SEG 29 (1979), 86. Henry (n. 14), 22, 31, 45, n. 2, 53, nn. 667 is
perhaps over-sceptical with regard to IG 22 2, given the parallel (noted by Henry) in SEG 29
(1979), 86. Further support for the use of a temporal clause is available in the deme honorific
decree of the fourth century IG 22 1210, lines 46:
|
|
.

TRAGIC HONOURS AND DEMOCRACY

21

We wait more than sixty years for the next examples of honours proclaimed at the
Dionysia.51 It is clear that this earlier group is special. While we can say nothing much
about the Boeotian father and son honoured in 403/2 B.C.,52 the other honorands form
a quite remarkable group: there is Thrasybulus (and probably Apollodorus), the
liberator of the Athenian dmos from tyranny; the saviour of Athenian manpower,
Epicerdes of Cyrene; and Euagoras king of Salamis, patron of Conons successful
arrival in Athens as a returning hero of renascent Athenian naval power all figures
key to the security and identity of the Athenian democracy across this critical
period.53 A recent study of the honours awarded the last on the list concludes that
they treat him like a destroyer of tyrants.54
Brief consideration of this group thus leaves no doubt about the importance of the
new practice of festival proclamation for Athenian democratic self-consciousness. Yet
it also requires that we nuance any interpretation of the announcement of honours at
the fifth-century Dionysia as above all a marker of Athenian democratic selfconfidence.55 While, as I have suggested, the granting of honours for Thrasybulus
does point to a new assertiveness of the returned democracy, the act itself hardly
obliterates the memory of what had given rise to it, no matter how anaemic,
euphemistic or evasive its language tried to be: in return for the good things he has
done [for the city] and the dmos of Athenians . The award of a crown to
Thrasybulus may speak a new confidence of the dmos in power, but that confidence
is at the same moment significantly tempered or tainted by a fragility and a sensitivity.
For the very act of crowning Thrasybulus inevitably recalls the moment at which the
shocking vulnerability of the dmos had been exposed. Read with an understanding
of its particular historical circumstances, such political self-confidence as we may find
in this practice is in fact not that far from the spirit of uncertainty or heavily qualified
optimism that tends to characterize the agency of the dmos or polis in tragedy itself.
To that extent, the contrast between tragic text and context may not be quite as sharp
as Goldhill presents it.56
The proclamation of an (earlier) award of a crown to Epicerdes of Cyrene, in
addition to another for a subsequent gift of emergency funds, is also especially
revealing in this regard. These large cash gifts to Athens had come from this foreign
euergets in an hour of extreme need, for the money was used to save the lives of
captured Athenian soldiers in Syracuse, in the darkest days the city had seen for a very

51 See Henry (n. 14), 312. The special honours announced (for the first time) at the
Panathenaea for the magnates of Bosporus date to 347/6 B.C.: IG 22 212, lines 2044, but we must
wait a further fourteen years after that for the next Dionysian example.
52 And see n. 50 above for reference to the doubts of Henry as to whether a crown is indeed
awarded, let alone proclaimed, in this fragmentary decree (IG 22 2 frag. b). M. Walbank, An
Athenian decree reconsidered: honours for Aristoxenos and another Boiotian, Echos du monde
classique / Classical Views 26 (1982), 25974, argues 403/2 B.C. is not possible as a date for this
inscription, and that the two fragments (a and b) do not derive from the same stl. He (tentatively) sees in frag. a a proxeny decree for the Boeotian exile Aristoxenus, dating it to 382/1 B.C.
Fragment b he argues is close in date, and perhaps honours with a crown and its proclamation
another Boeotian refugee who had provided funds to help the refugees from Boeotia.
53 Euagoras had been honoured (though without proclamation) eleven years earlier by the
Athenians, probably for securing their food supply: Lewis and Stroud (n. 15), 187: SEG 34 (1984),
24; Osborne (n. 8), 224 dates this decree to early 407 B.C.
54 Shear (n. 45), 1078.
55 Cf. Goldhill (n. 1), 623, 68.
56 Goldhill (n. 1), 68.

22

PETER WILSON

long time, with the threat of the extinction of democracy looming directly before it. In
saving the men, Epicerdes saved the city.57
The practice of proclaiming crowns to benefactors at the Dionysia thus simultaneously reveals the confidence and the fragility of the democracy, dependent as it was
on foreign and in many cases, extremely wealthy and powerful individuals, yet able,
in the very act of endowing them with such ostentatious honours, to assert and enact
its superior status in any relationship. The granting of citizenship as one of the
relevant ensemble of honours is an especially potent gesture. This highest award was
given infrequently, and can be viewed to some extent as a means of attempted
honorific control, of making its recipients at least notionally conform, in the act of
acceptance, to Athenian norms and standards.58
In recognizing this dynamic we may detect a further reason why it was the citys
premier agonistic festival that was the chosen site for this practice. For the very
business of awarding crowns to civic benefactors had the effect of assimilating their
actions to other forms of agonistic endeavour, of making doing good to the Athenian
dmos (euergetism) a contest like any other.59 The decision to timetable the proclamation of the winners of this special agn at the very moment of the citys most
prestigious cultural agn (tragedy) draws attention to the implicit assimilation of civic
euergetism to the long-familiar economy of the agonistic in the realm of performance
theatrical, musical and athletic.60 As in dramatic and musical contests, nonAthenians could and did always compete alongside Athenians in this event. But most
importantly of all, from the perspective of the Athenian dmos, this assimilation will
have served as a powerful form of honorific control. For it made of even the most
critical acts, vital to the very continuity of the democracy, performances to be judged
and appropriately awarded by the dmos.
BETWEEN OLIGARCHY AND DEMOCRACY:
THE DIONYSIA OF 410 AND 409 B.C.
The festival at which the assassins honours were to be proclaimed is that of 410/9 B.C.
This was surely an occasion of particularly strong emotions, in the aftermath of the
first oligarchic revolution of 411 B.C. and the restoration of democracy. But we should
first briefly consider the festival of the preceding year, 411/10 B.C. For there will have
been a number of unusual circumstances attending this earlier event.
The civic year had begun under the Archonship of Mnasilochus, the oligarchs man
who almost certainly went on to become one of the Thirty seven years later.61 Given
that it was the first practical duty of the Archon on entering office ([Arist.] Ath. Pol.
57 In Thucydides account (7.77.7) of the epigrammatic conclusion to the encouraging speech
to the soldiers in question delivered by their general Nicias, this disaster prompts a striking
absolute rhetorical identification between men and the city that draws on a topos of poetic
discourse:
. Alcaeus fr. 112, l. 19
(Lobel & Page); cf. Aesch. Pers. 349 and 352; Soph. OT 567.
58 This dynamic has been studied to good effect by A. Moreno, Feeding the Democracy: the
Athenian Grain Supply in the Fifth and Fourth Centuries BC (Oxford, 2007), esp. 261, 268 in his
analysis of the awards (though citizenship is not among them) granted by decree (IG 22 212) to
the kings of Bosporus in 347/6 B.C.
59 Thus Osborne (n. 15).
60 Cf. Burzachechi (n. 39), 109: ispirato certamente dalla consuetudine dell incoronazione
agonistica.
61 [Arist.] Ath. Pol. 33.1 with P. Rhodes, A Commentary on the Aristotelian Athenaion Politeia
(Oxford 1993 [1981]), 41011.

TRAGIC HONOURS AND DEMOCRACY

23

56.3), Mnasilochus is very likely to have appointed the chorgoi for the Dionysia, and
perhaps took further steps in the festivals planning. After the assassination of
Phrynichus in the autumn of 411 B.C., Theopompos replaced Mnasilochus as Archon,
but it is unlikely that he annulled the choregic arrangements that had been set in place.
This in any case heralded the start of the period of the moderate oligarchy of the
Five Thousand.62 The duration of that regime is unclear, but the restoration of full
democracy is generally thought to have taken place by March or April, following the
battle of Cyzicus.63 In other words, the Dionysia of 410 took place under the Five
Thousand, or just as their rule came to an end one reason, no doubt, why it took so
long for Erasinides & Co. to engineer the honours for Thrasybulus and his fellows.64
We happen to know something about one of the three chorgoi for the tragic
contest in this turbulent year. He was the extraordinary young man who commissioned Lysias twenty-first oration and we know that he spent half a talent in the
process (21.1).65 That sum is high and eloquent enough by itself. But, added to the list
of his other known services, whose total expenditure exceeds some ten talents, it
makes it as good as certain that he himself was from a family of compromised
oligarchs, desperate to buy his way out of trouble. Had he been appointed to the most
conspicuously prestigious choregic duty for tragedy by Mnasilochus as a promising
young sympathiser, only to find soon after that the need to present himself well
before the re-empowered dmos in that capacity had become all the more urgent? It is
no surprise that he describes himself as having come of age and performed this
chorgia in the archonship of Theopompus, (rather than Mnasilochus). This is
probably an expedient licence on his part. But even if he had been appointed by Theopompus, that meant appointment by the Archon of the Five Thousand, and probably
as we have just seen service at a festival with little by way of solid democratic
credentials.
The civic year of 410/9 B.C. was treated as a new democratic beginning.66 The very
first official political business of that year was the passing of the decree of Demophantus and its accompanying oath.67 This directed Athenian citizens to kill anyone
who attempted to overturn the democracy, granting them all legal and religious
impunity in doing so. It described citizens who responded to this call to defend the
democracy quite explicitly as tyrant slayers, in direct analogy with the tyrant-slayers

62 No democracy, pace G. de Ste Croix, The Constitution of the Five Thousand, Historia 5
(1956), 123.
63 D. Kagan, The Fall of the Athenian Empire (Ithaca 1987), 247. Others place the restoration
as late as June or as early as January: Meritt 1932 (n. 20), 10414.
64 Cf. Meiggs and Lewis (n. 9), 263; Gomme, Andrewes and Dover (n. 10), 310; Bearzot
(n. 10); and esp. Osborne (n. 8), 1820. Thucydides was unable or unwilling to name an assassin,
but rather attributed the murder to a nameless attendant (peripolos) of the Council Chamber
(8.92). And the evidence of Lysias shows that the matter was still far from settled in 400. The
facts were doubtless messy. Perhaps what the motion of Erasinides shows above all is the
ideological need, in the wake of this fatal breach of trust, for a single champion of democracy, a
latter-day tyrant-slayer with a recognizable identity (yet the long lapse of time taken, in addition
to the amendments of Diocles and Eudicus, demonstrate how difficult producing such a clear-cut
hero was).
65 He appears not to have been victorious: P. Wilson, Costing the Dionysia, in M. Revermann
and P. Wilson (edd.), Performance, Reception, Iconography: Studies in Honour of Oliver Taplin
(Oxford, 2008), 113.
66 Cf. esp. Andoc. 1.96.
67 Andoc. 1.97; M. Ostwald, The Athenian legislation against tyranny and subversion,
TAPhA 86 (1955), 10328.

24

PETER WILSON

Harmodius and Aristogeiton, and eliding in the process the oligarch and the tyrant.68
Anyone who died in the attempt, along with his children, was to receive the honours
accorded to them and their descendants.
It has generally been assumed (and occasionally argued) that these rewards were
restricted in their application to citizens only.69 But the language of the decree
describing the slayer of the would-be destroyer of democracy is broad, doubtless
deliberately so he who kills one who would do these things and he who conspires to
do so.70 While non-Athenians could hardly be bound by oath to protect Athenian
democracy, they could certainly be honoured for doing so on a par with citizens. Very
recent history will certainly have made the Athenians painfully aware of the value of
(presumably metic) non-citizens in the defence of their democracy.71
The accompanying oath is to be sworn by all the Athenians over adult offerings,
before the Dionysia (Andoc. 1. 98,
) the Dionysia in question being
that of 409 B.C. This phrase has generally been interpreted as a temporal index
stipulating a deadline before which all Athenians were required to have sworn the
oath.72 I suggest that it may have a more specific meaning, namely just prior to the
Dionysia and perhaps, more specifically, just prior to the tragic performances of
the Dionysia.73 It thus also serves the purpose of timetabling this important action
more precisely, as would be expected in a decree of this kind. In other words, we
should envisage the scene of the Athenians swearing the oath of Demophantus en
masse in the theatre, perhaps moments before the new and special honour of
68
69

See esp. Shear (n. 30), 152.

Shear (n. 30), 151 makes the case.

70
. The Athenians could
obviously legislate for non-citizens, including metics, within their territory, and in the absence of
further specification it is, I believe, preferable to assume that
includes citizens and
non-citizens alike. If however Shear (n. 30), 151 is right in arguing that non-citizens are not
included in these potential rewards, it makes the extremism of the oath all the more forceful. The
way it binds all citizens to become assassins under such terms takes shape in light of the fact that
the Athenians had evidently depended on foreigners to save them from oligarchy in 411.
71 The account recorded by Thucydides (8.92) is usually interpreted as suggesting that the
killer of Phrynichus was in fact an Athenian, for the unnamed peripolos is said to have had an
Argive accomplice, implying that the killer himself was Athenian: Gomme, Andrewes and Dover
(n. 10), 310. If there had been a potential Athenian candidate for the assassin, there must have
been powerful reasons why this account did not take root. One such reason would be that the
assassination of Phrynichus was in fact the product of in-fighting among the oligarchs rather
than noble resistance to their tyranny a version of events that the restored democracy would be
eager to obliterate. Bearzot (n. 10) makes some interesting arguments along these lines; cf. esp.
her conclusion at 298: La congiura contro Frinico nasce dallinterno della cercia degli oligarchi.
Note also the point made above about the controlled and controlling award of citizenship to
foreign benefactors, including Thrasybulus.
72 D. MacDowell, Andocides: On the Mysteries (Oxford, 1962), 136 inclines to the view that
the expression works only to stipulate a deadline for making the oath. But at the same time, he is
struck by the unusually lengthy period of grace allowed for the taking of the oath the surprisingly long one of about nine months.
73 See above p. 16 on the evidence for the placement of such activities at this moment within
the festival programme and below n. 77 for evidence that the phrase could be used with this more
specific meaning. I am delighted to see that Julia Shear (n. 30), 1538 recently arrived at the same
conclusion quite independently, although her interpretation differs from mine in locating the
oath-taking in the Agora rather than the theatre, and in placing the taking of the oath in the days
leading up to the festival proper. I refer to her excellent analysis, in which she further notes, for
instance, that the stipulation that all Athenians swear the oath by tribe and by deme is without
earlier parallel, and has a full discussion of the regular epigraphic usage of the preposition
in
the sense of just before which complements my own.

TRAGIC HONOURS AND DEMOCRACY

25

bestowing a gold crown before the same theatrical audience on one who had enacted
the spirit of Demophantus decree avant la lettre.74
There is no other example of
+
in the surviving corpus of Attic
inscriptions.75 And there is just a single instance of
+
. In the
honorific decree IG 22 351, lines 1920 (330/29 B.C.), Eudemus of Plataea is praised for
the following: he has donated a thousand (days of labour of ) yoke of oxen for the
construction of the stadium and the Panathenaic theatron, and has sent them all
before the Panathenaea, as he promised (
)76 While this may look rather more like an
instance of the deadline before which usage of
+ festival, it also demonstrates
the special purposefulness that motivates the selection of the festival mentioned in
such phrases for there is no doubt that the oxen provided by Eudemus were to be
employed for heavy haulage and construction-work on the stadium and theatron
needing to be readied for use at the next Panathenaea.
Moreover, the epigraphic examples cited by Shear (n. 30), 1556 in reference to
in the
events or festivals other than the Dionysia demonstrate more fully that
sense of just before was a familiar usage. Among these, the direction to proclaim
honours awarded by the Mesogeioi is especially illuminating. These are to be
announced in the sanctuary of [Heracles at the festival], just before the sacrifice
|
|
) IG 22
(
1244, lines 35.
In addition, the single literary example of
generated by a
textual search of the TLG confirms the view that in the decree of Demophantus the
phrase should mean just prior to the Dionysia. This is the account given by Aelian
of the young Plato, on his way to the theatre to compete with a tragic tetralogy that
he had composed, only to hear and forever be distracted by Socrates:
passing by just before the Dionysia he
heard Socrates , VH 2.29.77
74 The interpretation of the phrase I propose (all the Athenians are to swear it just prior to
the Dionysia) effectively includes the purpose of the more general interpretation as deadline (all
the Athenians are to swear it before the date of the Dionysia), on the assumption (easy enough in
terms of ordinary Athenian ideological identifications: see esp. Dem. 21.18 with P. Wilson,
Demosthenes 21, Against Meidias: democratic abuse, PCPS 37 [1991], 164, 1878; Shear [n. 30],
156) that all the Athenians are present for the Dionysia and swear the oath before it begins in
earnest. If it did take place in the theatre using an altar in the orchstra, or perhaps by filing into
the sanctuary the requirement in Andoc. 1.97 that the oath be sworn
has intriguing and hitherto neglected consequences for the (or at least one possible)
disposition of the audience.
75 Nor is the Dionysia used in the way the Panathenaic quadrennium functions, to divide civic
time into blocks
: e.g. Ath. Pol. 43.1; IG 13 52, A lines 278, b.
line 28; IG 13 292, line 3; IG 13 300, line 2; IG 13 317, lines 12, etc.
76 On this inscription see Lambert (n. 42). The recognition that
= day equivalence goes
back to A. Wilhelm, Neue Beitrge zur griechischen Inschriftenkunde, SAWW 166 (1911),
163, at 49. This amounts to a commitment to supply the necessary thousand days of labour
before the festival. Given that Eudemus offer was made (according to the decree) on or near 11th
Thargelion, the next Panathenaea was only about eight or nine weeks away. The thousand days of
promised labour are therefore, I suggest, very likely to represent something like fifty calendar
days of work with twenty pair of oxen. My thanks to William Slater for discussion of this point.
77 This use of
(
)
also shows that the phrase could mean before the theatrical performances of the Dionysia rather than, as strict logic might seem to require, before the
start of the festival of Dionysus. If the latter were the required meaning of
in
the decree of Demophantus, as Shear (n. 30), 157 suggests, it points to Elaphebolion 9 as the day

26

PETER WILSON

It is thus highly likely that the collective Athenian enunciation of the important
oath of Demophantus against anti-democrats is to be placed among the very special
set of pre-tragic ceremonies that took place at the City Dionysia of 409 B.C. Moreover,
the evidence of Aristophanes Birds (10745) suggests that in 414 B.C. a practice had
long been in place for the announcement in the theatre, prior to the performance of
drama, of awards for the murder of aspiring tyrants.78 If Demophantus oath was
indeed proclaimed collectively by the Athenians in the theatre just prior to the
Dionysia of 409 B.C., it presents itself as an updating of this older practice in light of
recent experience.79 So, whether we are to imagine the new oath of Demophantus or
the old announcement of awards for the slayers of would-be tyrants as the accompaniment to the proclamation of the gold crown for Thrasybulus, the Dionysia of 409
B.C. saw an extraordinary agglomeration of rituals shoring up and shaping democratic
ideology.80
Indeed, as I have already indicated, the existence, from perhaps as early as the
beginning of the fifth century, of this proclamation at the Dionysia of rewards for
tyrant-slayers forms a neglected element in the portfolio of evidence for the politics of
the theatre, and more particularly for discussion of the relation between tragedy and
its festival frame.81 In instituting the practice of awarding and proclaiming crowns for
benefactors at the tragic contest of the City Dionysia in the tumultuous final decade

for the oath, since the Dionysia officially began on or around Elaphebolion 10 with the pomp,
with the dramatic contests starting on the 11th: Csapo and Slater (n. 36), 1067. Elaphebolion 9
is a day for which meetings of the Assembly are known: J. Mikalson, The Sacred and Civil
Calendar of the Athenian Year (Princeton, 1975), 12330, 137. It is true that much the more usual
way to refer to the moment of the pre-performance ceremonials is by use of a dative (
,
, etc.). I would therefore not insist that
has precisely
the same meaning here. It is possible that the variation is rather intended to place the hugely
important activity of oath-swearing at the very beginning of the sequence of various pre-performance ceremonials. But even if the phrase was not intended to tie the oath-taking physically to
the theatre, the underlying politics of its association with the Dionysia remain unchanged.
78
/
/
. N. Dunbar, Aristophanes: Birds (Oxford, 1995),
581 especially on the question as to whether this was confined to the first day of the Dionysia or
took place on each day. See further K. Raaflaub, Stick and glue: the function of tyranny in
fifth-century Athenian democracy, in K. Morgan (ed.), Popular Tyranny: Sovereignty and its
Discontents in Ancient Greece (Austin, 2003), 5993, at 6970, who accepts the evidence for the
practice and Seaford (n. 2), 346.
79 Cf. Raaflaub (n. 78), 70.
80 The combination of a decree about tyrant-slaying and an ideologically charged act of
crowning that we find at the Dionysia of 409 B.C. reappears in an intriguing manner in the law of
Eucrates of 337/6 B.C. (SEG 12 [1955], 87), which clearly looks back to Demophantus provisions.
Whereas in 409 B.C. the decree and crowning were separate acts, the two are collapsed into one on
the stone of the later law, as relief carving (Dmokratia crowns Dmos) and accompanying law. I
thank Alastair Blanshard for this observation and refer to his fine study of the Eucrates
monument: A. Blanshard, Depicting democracy: an exploration of art and text in the law of
Eukrates, JHS 124 (2004), 115.
81 It is difficult to imagine an historical context other than that around 500 B.C. which would
have given rise to the institution of the proclamation in favour of tyrant-slayers at the Dionysia.
The lines from the Birds play on the idea that tyrants are a thing of the distant past. This might in
itself be taken to suggest an awareness that the measure was of very long standing. More salient,
however, is the fact that it draws its humour from a deliberate misrecognition of the way the
Athenians re-figured the tyrant over the course of the fifth century. While affecting to present
tyrants as a thing of the distant past, it draws its impact from a no doubt keen awareness that, in
414 B.C., in the wake of the mutilation of the Herms and profanation of the Mysteries, and all
that those acts portended, fear of tyranny was a reality in the city once more.

TRAGIC HONOURS AND DEMOCRACY

27

of the fifth century, in close association with the decree and oath of Demophantus,
the Athenians were thus activating the perhaps largely dormant potential of a ritual
proclamation the democratic orientation of which is beyond doubt. That ritual, the
mentalities that underpinned it, and the ease with which it was renovated and
rearticulated for the special conditions that afflicted the democratic city in 410 B.C.
show that the tragic contest of the Dionysia had been the natural home for such
democratic expression for the whole period from which our surviving dramas derive,
and that we are surely right to continue, with Goldhill, to think of the institutional
framework within which drama was performed at Athens as fundamentally democratic, rather than simply a polis framework.82
The theatre came to this role in 409 B.C. with its own history. Indeed, we might say
that the theatre itself had been a protagonist in recent political history, and that the
attention devoted to the site in the aftermath of 411 B.C. to some extent reflects and
refers to the role it had played. For Thucydides makes it clear that the two principal
theatrical spaces of Athens that at Munichia in Piraeus and the urban theatre by the
Acropolis were the chosen sites of democratic resurgence and formal decisionmaking during the short period of oligarchic rule. In an important sense, in this
period the theatre served as a shadow-site of the abolished democratic Assembly. The
one in the Piraeus was where the first concerted action of the dmos under arms took
place, in the wake of the murder of Phrynichus. The hoplites stockpiled weapons
there and held an assembly in which they resolved to march on the city. Once there,
the theatre of Dionysus on the slopes of the Acropolis became the site chosen for the
meeting of the hoplite assembly on a fixed day, to discuss homonoia. That meeting
never took place, but the plan suggests a propensity on the part of the anti-oligarchs
to turn to that space in the hard journey towards political resolution.83
The rich young speaker of Lysias 21, of uncertain political allegiance, offers a
point of striking continuity between the two successive Dionysia following the oligarchic revolution of 411 B.C. For he appears once again at the festival of 409 B.C.,
where the tragic contest was certainly preceded by the announcement of honours for
Thrasybulus and his fellows, and the placing of a gold crown on the head of an
Athenian euergete for the very first time and one who had in effect liberated the
dmos. And by a rare and happy coincidence of available evidence, we know that the
winner of this tragic agn was none other than Sophocles, with the set of dramas
that included the Philoctetes.84 It is worth recalling the poets own recent very direct
involvement in the events that led to the abolition of full democracy, as one of the
probouloi who voted for the Four Hundred; and noting that this evidently did not
tarnish his reputation with the Archon Glaucippus, or with the judges of the tragic
agn of that year.85
82

Rhodes (n. 2), 119.

Thuc. 8.93.1, 8.93.3, 8.94.1. See Gomme, Andrewes and Dover (n. 10), 316; N. Loraux, La
Voix endeuille: essai sur la tragdie grecque (Paris, 1999).
84 Prose hypothesis Soph. Phil. The proximity of these major theatrical and political events at
the Dionysia of 409 B.C. has never, to my knowledge, been noted by Sophoclean critics or historians of the theatre (however see now Shear [n. 30]). There is for instance nothing in C. Greengard,
Theatre in Crisis: Sophocles Reconstruction of Genre and Politics in Philoctetes (Amsterdam,
1987), a work which argues for a much closer relation of the drama to its historical setting than is
normally accorded these timeless dramas. (p. 8).
85 Sophocles as proboulos: Arist. Rh. 3, 18, 1419a25 = TrGF 4 T Gd 27. Cf. M. Jameson,
Politics and the Philoctetes, CPh 51 (1956), 21727, esp. 21718. Without entering into the
debate as to whether the praise of Colonus in the OC (668719) the place where a meeting of
83

28

PETER WILSON

The choice of role made by the young liturgist of Lysias 21 at the Dionysia of 409
is particularly interesting. This time he served as chorgos for the mens chorus of
his tribe (Lys. 21.2), and this time he spent the vast sum of 5,000 drachmas
including the erection of the tripod, this last phrase indicating that his chorus was
successful in the agn. He will certainly have volunteered for this task, as his tragic
chorgia in the immediately preceding year gave him automatic exemption from any
such obligation.86 His decision to support his tribe, and to do so in so lavish a fashion,
may have been motivated in part at least by a wish to identify himself with the more
democratically configured performance category of the mens chorus. It could thus
be seen as an act of symbolic identification with the restored regime at a time when
reprisals against those who had been implicated with the rule of the Four Hundred
were widening in scope.87 The sheer scale of his expenditure is striking. There are
many practical reasons that might make one expect a larger outlay on the mens
chorus than tragedy or comedy the fact that the former had fifty to tragedys fifteen
singer-dancers prominent among them. But in this case we may suspect a wish, or
indeed a need, on the part of this chorgos, to be able to say that his expenditure on
the performance with a greater collective affiliation and at a festival with no shadow
cast over its organization had far outweighed that on tragedy in the previous year.
This has the air of a compromised individual buying the charisma and charis of
victory in the newly democratically sanctioned festival at any cost.
As victorious chorgos for the mens chorus he too will have been crowned with
ivy rather than gold at the end of the proceedings in the theatre that began with
another, very special crowning. How did this rich young man, seated in a prominent
proedric seat as one of the years chorgoi, hear the announcement of these honours
for the democratic assassins at the Dionysia, as he saw the gold crown placed on the
head of Thrasybulus, euergets of the Athenian dmos? The internal dynamic of this
message was surely intended above all for him and his kind a wealthy citizen who,
while perhaps no fervent oligarch (witness his service at Aegospotami and his
employment of Lysias, no friend of tyrants), all the same found no difficulty being
complicit with an oligarchic regime (as his prominence under the Thirty, and his very
presence in the city in Hecatombaion 403 B.C. show). The question of whether the
chorgoi for the Dionysia of 410 B.C. had indeed been appointed by the oligarchic
Archon Mnasilochus becomes highly charged when we reflect on the fact that the
oath of Demophantus explicitly included among those who were thereafter to be
piously assassinated anyone who holds office after the democracy has been overB.C.

the Assembly under Peisander was held reflects a partisan or apologetic oligarchic view on the
part of the poet (see e.g. J. Wilson, The Hero and the City: An Interpretation of Sophocles
Oedipus at Colonus [Michigan, 1997], 1989; L. Edmunds, Theatrical Space and Historical
Space in Sophocles Oedipus at Colonus [Lanham/London, 1996], 91110), it is worth observing
the contrast between that Assembly, held in the sanctuary of Poseidon Hippius at Colonus, and
the ones held in the theatre(s) of Dionysus during the oligarchy, but also thereafter, to foreswear
oligarchs and tyrants. For further bibliography and a balanced discussion of political readings of
Sophocles OC see A. Markantonatos, Oedipus at Colonus: Sophocles, Athens, and the World
(Berlin/New York, 2007), 1215.
86 Dem. 20.8; 21.155; 50.9; Ath. Pol. 56.3. Isoc. 7.38 implies that an exemption period of two
years may have operated in an earlier time: MacDowell (n. 23), 127, n. 3. It is not inconceivable
that choregic appointments made by Mnasilochus for 410 B.C. were subsequently deemed
formally invalid, at least in respect of attracting exemption to their holders. Such a deduction
should not, however, be made solely on the basis of Lys. 21, for its speaker repeatedly ignored the
availability to him of liturgical exemption over the course of several years.
87 Wilson (n. 2), 90.

TRAGIC HONOURS AND DEMOCRACY

29

thrown (
Andoc. 1.97). Chorgoi were not archontes but they were sufficiently indistinct from
them to prompt a discussion of the question from Aristotle (Politics 1299a1520).88
And the chorgoi of the Dionysia in 410 B.C. had, strictly speaking, probably only
served under a narrowly oligarchic regime for a brief period and under the Five
Thousand for a further four or five months. But the situation was doubtless close
enough to make this an uncomfortable moment for this young man.
The decision to present Thrasybulus before the panhellenic audience of the Dionysia
at the tragic agn of 409 B.C. can hardly have been merely a matter of convenience. If,
as is likely, IG 13 102 represents the first use of the festival in this way, it is important
to reflect on the fact that the practice was designed for and initiated in this context
the desire to honour the first historical saviour of the developed democracy.
But it is also important to recognize that this innovation itself represents a
continuity, the development of a long-familiar association between the tragic contest
of the Dionysia and the defining co-ordinates of democracy. This is not the first, nor
will it be the last, example of Dionysiac ritual moulded to meet the needs of a society
articulating its sense of political liberty.89
University of Sydney

88

PETER WILSON
peter.wilson@usyd.edu.au

Wilson (n. 54), 16970.

See Connor (n. 2). Another Athenian example: Philippides, comic poet and statesman, as
agonothete for the City Dionysia in 287 B.C. was the first to arrange an extra agn for Demeter
and Kore as a reminder of the [liberation] of the dmos (IG 22 657, lines 435); and cf. also the
Eretrian decree of around 309 B.C. (12. 9 192 = LSCGSuppl. 46) which prescribes inter alia the
wearing of ivy crowns at the pomp of Dionysus by all, citizens and metics, as a memorial of the
day when during the pomp of Dionysus the [Macedonian] garrison departed, the dmos was
liberated and reintroduced democracy (lines 35) with A. Jacottet, Le lierre de la libert,
ZPE 90 (1990), 1506; P. Wilson, The politics of dance: dithyrambic contest and social order in
Greece, in D. Phillips and D. Pritchard (edd.), Sport and Festival in the Ancient Greek World
(Swansea/London, 2003), 16598, at 1801.
89

Classical Quarterly 59.1 3045 (2009) Printed in Great Britain

doi:10.1017/S00098388090000032

30
A N A R IS TO PH A N IC S LAVE: DANIEL
PEACE 8191126
WALIN

AN ARISTOPHANIC SLAVE: PEACE 8191126

When discussing Aristophanic slaves, it is usual to observe that Xanthias in Frogs and
Carion in Wealth are exceptional in the extant plays, and that they occupy a place
transitional between the slaves of Old Comedy and those who would develop later.
While conceding that there is a fundamentally new (as far as we can tell) approach to
the characterization of the slave in Aristophanes treatment of Xanthias in 405 B.C.E.,
I would like to explore in depth one of the forerunners of these characters, the
in Peace 8191126, who appeared on the stage sixteen years prior in 421 B.C.E.
and shares many characteristics with Xanthias. I will examine how he often acts not as
a slave but as an equal of his master, how he tries to attract attention while on stage by
volunteering his humour, his preoccupation with the sexual, his possible use of asides,
the dynamics of his unruliness and why his master never threatens or punishes him,
and finally what it means that the audience never learns his name.
Peace begins with a pair of slaves kneading cakes for a giant dung beetle,1 on which
the hero of the play, Trygaeus, plans to fly up to heaven in a parody of the flight of
Bellerophon on Pegasus (cf. Pax 76, 1359) humorously influenced by the Aesopic
fable of the beetle and the eagle (cf. Pax 12934; Aesop, Fab. 3) to ask Zeus what he
intends to do with the Greeks. When the hero arrives, however, he learns from Hermes
that the Olympian gods, frustrated because the Greeks refuse to accept Peace, have
left Greece at the mercy of War, who intends to grind up all the cities of Greece in his
mortar but is prevented by lack of a pestle.2 This affords Trygaeus the opportunity of
summoning a panhellenic chorus to help him rescue Peace from the cave where she is
imprisoned. Trygaeus must make an ally of Hermes by bribing him with a golden cup
and promising all the rites formerly given to the other gods. After a great deal of
effort, the chorus rescues Peace and her handmaidens, Opora (Harvest), who is to be
. As
the bride of Trygaeus, and Theoria (Festival), who will be given to the
Trygaeus returns to earth with these two women, he encounters one of his slaves, who
operates for the moment as a straight man for his master by asking him questions
about his travels that provide an opportunity for a series of jokes (82441).3
1 It is possible that one of these slaves is the same character as the
at 8191126.
Sommerstein (1985), 172 and Olson (1998), xliii, 231 suspect that this character may be identical
with the second slave from the prologue (i.e. the more dominant one, who remains on stage to
speak the prologue when the first has gone inside to feed the beetle, and who engages Trygaeus in
dialogue as he is flying away to heaven). Olson goes so far as to label as
both the
prologue slave and the one at 8191126. If we accept this identification (and there is no compelling reason why we should not), it only magnifies the role of this slave, which would strengthen
the arguments I make here. It would also give the slave a more metatheatrical role, because he
would speak to or about the audience not only at his appearance late in the play (as at 8834 and
96373) but also at the beginning. An increased metatheatrical role, in turn, would make it more
likely that he does speak in asides (see below). The audience would have known one way or the
other by the mask and dress of the character(s).
2 The lost pestles are Cleon and Brasidas, the most ardent war-hawks of each side. Both
perished at Amphipolis in 422 B.C.E. Cf. Thuc. 5.10.
3 Trygaeus and this slave are partners in humour. They take turns performing the role of
straight man and telling the jokes, so that now one, now the other is dominant on stage (inasmuch
as control of the jokes is control of the play).

31

DANIEL WALIN

Trygaeus abruptly ends the string of jokes about his travels by ordering his slave to
take Opora inside and prepare her ritual bath and the wedding bed, while he himself
(which despite this line he does not do until after his
takes Theoria to the
slave has returned). Apparently noticing the women for the first time, the slave asks
his master where he got them. When he finds out, he expresses some surprise that
/
even the gods keep prostitutes (
, 8489) a comment that we might expect to
offend his master, who after all is about to marry one of these girls.4 But Trygaeus
takes the comment in his stride and even seems to affirm that these symbolic women
are prostitutes (
, 850).5 This is the first of
many examples of the tolerance and leniency of Trygaeus as a master, a trait that
arguably explains and/or induces the presumption of his slave.
The slave next asks a seemingly innocuous question: should he give Opora anything
to eat? With his response Trygaeus (whether intentionally or not) sets up a joke for his
slave:
{

{Sl.}
{Tr.}
{Sl.}

(Pax 8515)

Come on, lets go. Tell me, should I give her something to eat up?
No. She wont want bread or barley-cake, when with the gods above
shes used to licking ambrosia.
Shell have to prepare herself to do some licking here, too!

The joke here revolves around the range of meanings of

(to lick), which can
apparently indicate either eating or fellatio.7 There are several important things to
note here. First, the slave has for a moment become the primary speaker of a joke; his
masters role of setting up the joke is secondary. We shall see more of this later.
Second, the slave has spoken what may be an aside at 855.8 Immediately after 855 the
4 Stefanis (1980), 157 classifies the slaves comment here with other such passages under the
heading the slave as appraiser of gods and men (
). The
phrasing of the Greek does suggest the translation keep prostitutes, if not something more
suggestive of an analogy with the keeping of cattle. One
various kinds of domesticated
animal, though the metaphorical use of the word is by no means confined to this passage. The
chorus of old men in Lys. uses the term both of ordinary women (i.e. wives) at 25665 and of
their slaves (
, which could refer more broadly to their dependents in general) and children
(

) at 12034.
5 The whoredom or adultery of female symbolic figures seems to be conventional in extant
comedy. For another example in this play, cf. Pax 97890 (discussed in depth below). Whether the
nude female mute in Old Comedy was played by real naked (or scantily clad) prostitutes or by
padded male actors is an inveterate controversy, for a useful discussion of which see B. Zweig,
The mute nude female characters in Aristophanes plays, in A. Richlin (ed.), Pornography and
Representation in Greece and Rome (Oxford, 1992), 7389. If Opora and Theoria are being represented by actual courtesans, the slaves comment is quite metatheatrical (though in any case the
slaves comparison of the goddess Peace to an adulterous woman in his prayer at 97890 is not
metatheatre, for she would have been represented by a statue).
6 I print the Greek text of Pax as in Olson (1998). Translations are my own.
7 Cf. Olson (1998), 235.
8
K.J. Dover, Aristophanes Frogs (Oxford, 1993), 445 rightly notes Xanthias ability to speak
in asides, an ability which helps to characterize him as an entity independent of (even better than)

32

AN ARISTOPHANIC SLAVE: PEACE 8191126

slave and Opora leave the stage, and a choral ode begins, so that there is no verbal
response to his remark. The key question is whom the audience is meant to envisage
as the recipient of Oporas attentions. If it is the master, he may have no reason to
react to the statement although we should still take note of the slaves interest in the
details of the sexual relationship between his master and Opora and the masters
acceptance of that interest but if it is the slave, we should perhaps expect some
reaction from Trygaeus (provided that he hears). Action on stage could have made
clear to the audience both whom the slave intended and any nonverbal reaction on the
part of Trygaeus.9 Because we have only a written text, we cannot know for sure how
the scene played out. But I think that the lasciviousness of this slave as expressed
elsewhere his willingness openly to scrutinize and fondle the women his master
brings home makes it likely that he does mean himself, especially when he speaks
this line as he and Opora enter the house, away from the eyes of Trygaeus.10 He could
also be implying that Opora will bestow her favours on both of them, which would be
in the spirit of their shared sexual experience of Theoria, alleged by the slave at Peace
8716 (see below).
It is certainly true that Opora and Theoria are broadly conceived of in this play as
respectively the private and public benefits of peace, the former the prize of Trygaeus,
. Hence it is possible to suspect an interpretation in which
the latter that of the
the sexual benefits of Opora do not belong solely to Trygaeus on the grounds that it
are not
would undermine this dichotomy. But I argue that moments of
subject to the generalizations that apply to the play as a whole. In such a moment
humorous effect is paramount, and disruption or subversion of what the play is
otherwise trying to accomplish should not be unexpected. After all, Aristophanic
comedy is full of gag-scenes that go nowhere and only serve the exigencies of humour.
So if the slave does imply that he wants to or will have sex with Opora, he does so
because (presumably) it is funny. This does not need to affect much our interpretation
of what Opora and Theoria mean for the play as a whole.
In his recent book Humour, Obscenity, and Aristophanes James Robson attempts to
understand better Aristophanic humour in part through the application of a theory
of the frame, by which he means a generalized context of experience which we
subconsciously organize by a system of unwritten rules.11 In the detailed analysis of
Peace 819921, which comprises his final chapter, he observes the substantial interval
between the (probable) appearance on stage of Opora and Theoria with Trygaeus at
819 and the first reference to them in the dialogue at 842, noting that elsewhere in
Aristophanes male characters respond lasciviously to the presence of mute nude
female characters. He argues that the mere presence of such characters establishes a
his master. He observes three lines (88, 107 and 115) where we as readers can discern with confidence that Xanthias is speaking in asides. Other lines (334, 41, 51, 308, 311) may or may not be
asides.
9 I think the probability that Trygaeus hears but reacts only non-verbally is slim. Certainly we
might expect that, if this were the case, he would make some comment later on, for instance when
the slave admires Oporas rump at 868.
10 For other instances of the lasciviousness of this
, cf. Pax 86880, 8913 and the
discussions below. But he and his master are a team in dirty jokes as much as in other things (Pax
874), such that this characteristic cannot really serve to distinguish him from his master (cf. Pax
8845, 894904).
11 J. Robson, Humour, Obscenity, and Aristophanes (Tbingen, 2006), 17. He illuminates the
concept by the example of the frame buying a train ticket, in which certain linguistic sequences
(e.g. What time does the train leave? or How much does it cost?) are expected while others are
not (e.g. a discussion of aubergine farming in Malaysia).

33

DANIEL WALIN

certain frame in which the audience is led to expect lascivious comments, so that the
whole time the women are on stage the audience is unusually attuned to the possibility
of double entendre. It is concluded that the postponement of true obscenity to the
, 855) tantalizes the audilast line of the scene (
ence.12 This argument lends weight to the slave as a character, for it posits an audience
which anticipates sex jokes the entire time Trygaeus is responsible for the humour, all
the while not knowing that what they are really waiting for is a transfer of comedic
agency from the master to the slave, for it is in exploiting the comic potential of Opora
and Theoria that the slave will find his first outlet for a joke of his own. As we shall
see, he continues to find in these women (and later, as I argue, in their mistress) a rich
source of material.
If the slave communicates a desire to have oral sex with Opora to the audience in an
aside, and does so immediately before the two of them leave the stage together, so that
it is unclear whether this desire will come to fruition during the time allotted for her
bath, it contributes considerably to the slaves power over the drama and displays a
rivalry with or dominance over his master that approaches that of Xanthias in Frogs.
In this case the choral ode which immediately follows may admit of an ironic interpretation: the chorus says that Trygaeus is faring well as far as can be seen (
-/
/
, 8568) and emphasizes
his old age twice (
, 8567;
, 860),13 thus playing on the possibility that
he is being cuckolded as they speak. When Trygaeus imagines himself as a shining
bridegroom fondling Oporas breasts (85964), the audience may think of the slave
really doing offstage what his master is then imagining; and again there would be an
element of irony when Trygaeus says that he has saved Greece so that everybody can
/
screw and sleep safely (
, 8667).14
What can be said with certainty is that the slave both leaves (855) and reenters (868)
the stage with a sexual remark about Opora, his masters intended. He returns with
the following locution:
{

15

(Pax 86870)
{Sl.}

The girls been bathed, and her arse is doing fine.

The cakes been baked, the sesame is being shaped,
and all the rest. But she does need a dick!

The slaves frank admiration of Oporas arse, unlike his reference to fellatio at 855, is
probably not an aside, for it is part of the same line in which he reports to his master
that his order that she be bathed has been obeyed. We must assume that Trygaeus
hears this remark, but for whatever reason he does not seem to mind. The remark
about the need for a penis (870) does not in itself refer to the penis of any one person
in particular, so that we may expect that Trygaeus will take it as applying to himself,

12

Ibid. 15762.
Old men with young, attractive brides are natural targets for cuckoldry.
My anonymous reader makes a legitimate objection: the tone of parallel choral odes in
Aristophanes is generally similar, and there are clearly no such shenanigans going on at 90921.
But a general rule is precisely that and may be violated on occasion.
15 Apparently
here is
for
Olson (1998), 237.
13
14

34

AN ARISTOPHANIC SLAVE: PEACE 8191126

while the audiences interpretation would have been influenced by whatever clues in
staging there may or may not have been.
Trygaeus lack of reaction to the lascivious comments of his slave could be
explained by an excessive (to the Athenian audience) degree of leniency and a way of
regarding the masterslave relationship as a partnership, exemplified in the fact that
the two have shared a sex partner before in Theoria:
{

{
{

}
}

16

(Pax 8716)
{Tr.}
{Sl.}
{Tr.}
{Sl.}

Come then, lets hurry up, you and I, and give Theoria here to the
Council.
Whos this? What are you saying? This is Theoria, whom we used to bang
to Brauron when we were a bit drunk?
It sure is and catching her was quite a task!
Master, what a quadrennial arse she has!

Trygaeus clearly hears the slave, for he responds at 875. If we read the passage literally,
Trygaeus and his slave were banging Theoria to Brauron; the easiest interpretation
of this phrasing, in my opinion, is that the force of their sexual activity was so great
as a (potentially)
that Theoria ended up in eastern Attica.17 RV recognize
sexual word, glossing
as
. There is also a reference to the quadrennial festival of Artemis at Brauron, which may have been a ritual
of maturation for Athenian girls and therefore perhaps a prime opportunity for sex.18
As Sommerstein and others have noted, the real meaning of this passage seems to be
that Trygaeus and his slave used to frequent this festival, and that their experience
there was marked by sexual activity.19 The slaves comment exploits the difference
between the female character on stage who though mute is real enough as far as
characters in a comedy go and the travel to and attendance at festivals that she
is
symbolizes. Getting laid at festivals becomes getting laid with Festival.
apparently
for some word such as
or
.20 The
change to
stresses sex, (presumably) one of the most pleasant aspects of the
festival, and allows for a joke based on the double identity of Theoria.21 The slaves
16 The anonymous reader suggests that Trygaeus means Theoria is an exclusive prostitute. I
think it is ambiguous whether Trygaeus (1) means that his struggles to free Peace and her
handmaidens were arduous or (2) refers in some way to what has just been said by the slave. In the
latter case there are several possibilities: Theoria could be envisaged as an exclusive prostitute, a
rape victim, or a willing sexual partner who was scarcely taken because of sexual impotence
induced by drunkenness (
)
17 For this use of
, cf. Olson (1998), 238; J. Henderson, The Maculate Muse: Obscene
Language in Attic Comedy2 (Oxford, 1991), 171 (308). Henderson discusses the use of violent
words as sexual slang under the heading Hitting and Piercing, 1703 (298316).
18 Olson (1998), 238.
19 Sommerstein (1985), 174.
20 Sommerstein (1985), 174; Olson (1998), 238.
21 Platnauer (1964), 142 denies (against
RV and the more recent editions) that
has a
sexual sense, suggesting instead Kocks emendation
, which (I think) is more likely the
expected word displaced by a
joke than what one should actually read here.

DANIEL WALIN

35

line about her quadrennial arse is fairly typical for Aristophanic comedy; the female
body on stage often becomes a prop for jokes.22 I think that the metatheatre here
should be a warning not to read too much into the fact that, according to a strict
interpretation of the text, the master and slave have had their way with the same
at 873 is ambiguous: the
woman, either simultaneously or in close succession.23
slave, as the speaker, is necessarily included, but he could make a sweeping gesture to
include not only his master and himself but also the entire audience, and in so doing
he would become for that moment not a slave in comedy but an Athenian citizen, an
actor appealing to the desirability of the communal experience of festivals as an
incentive for peace.24 But however one reads it, the slave is playing a dominant role in
the humour at this point.
The slave continues to use Theoria as a prop for humour, not only with words but
even with actions. As his master is trying to decide to whom out of the whole audience
he should entrust Theoria and implying that whoever can be trusted to guard her for
is
(8778) the slave is proving his lack of
by groping
the
her:
{
{

}
}
(Pax 87980)

{Tr.}
{Sl.}

You! Just what are you circumscribing?

Umm hmm for the Isthmian Games Im claiming a campground for
my dick!

The groping prompts a question from Trygaeus, which in turn sets up a joke for the
not necessarily the most natural word for
slave (the use of the verb
Trygaeus in this context is clearly for the sake of the joke that follows).25 The
reference to the Isthmian Games particularly in a year (421 B.C.E.) in which there
were none is a joke; as Sommerstein puts it, the isthmus [the slave] is thinking of
is the narrow strip of territory between the two broad expanses of [Theorias]
, 880) in which the slave plans to rest his penis is obvious.
thighs.26 The tent (
Trygaeus, however, ignores these remarks as he has done, in what begins to be a
pattern, with 86870 and 876.27
Possible explanations for Trygaeus lack of reaction to his slaves lascivious
comments so far have been (1) that the master is uncommonly lenient (which he
certainly is, for he never threatens violence to his slave and does without doubt
sometimes treat him, if only for a moment, as a partner) and (2) that the meta-

22 For other examples (by no means a comprehensive list), cf. Pax 87980, 88793; Lys. 879,
912 and 111288. The first two examples from Lys. are interesting in that it is the women who
use the bodies of other women for sexual humour. But the male body is also the object of
humour, and it is perhaps safest to make no distinction and say merely that it is quite common in
comedy to point at someone elses body parts and make humorous comments.
23 But the slave and Trygaeus may also be imagined as sharing the goddess Peace as a mistress
at 97890.
24 Aristophanes often promotes peace by emphasizing the many good things that come from it
and reminding the audience of the hardships of war.
25 Cf. Olson (1998), 238.
26 Sommerstein (1985), 175. For this sense of the word isthmus he refers the reader to Aesch.
fr. 17.2931M = F 78a 2931 Radt, from a satyr play called Theoroi or Isthmiastai.
27 855 too, if we assume that it is not an aside and that Trygaeus hears it, is probably ignored.

36

AN ARISTOPHANIC SLAVE: PEACE 8191126

theatrical nature of some of the slaves comments (for example, the ambiguous
treatment of Theoria as both festival and woman) inclines both characters and
audience to consider them apart from the flow of the play, so that the masters
response (essentially yes, this is that same girl we used to screw) is partly the response
of Trygaeus, partly that of the actor who plays him.28 But there is another consideration best illustrated by adducing a passage from Wasps:
{

}
(Vesp. 5002)

{Xa.}

Yesterday my whore got mad at me too, when I came in to her at noon,

because I ordered her to ride me. She asked if I was setting up a tyranny
in the style of Hippias!

In what immediately precedes, Bdelucleon remarks on how easily the term tyrant is
bandied about, using as an example the talk in the market. It is his slave that
introduces sexuality to the passage in these three lines. Though the remarks of the
slave are not necessarily an aside three lines do seem rather long for that they are at
any rate ignored. Perhaps one of the roles of the slave character visible here and
throughout the part of Peace in which we are engaged is to be a source for jokes that
are funny enough, but do not really advance the plot or the main characters purpose
at the time, so that it is better for them to be spoken by someone other than him or
her. If the Athenian audience expects a slave to be making occasional wisecracks
certainly it would later, when it had become accustomed to such figures as Xanthias
and Carion, but we are speaking of the early comedies then it becomes too much a
drag on the plot (and not very funny) for the comments of the slave always to require
a response. In this view it would have become a convention that slaves do not always
get the rebuke they might have received in real life (when is Old Comedy ever truly
realistic?), and when they are rebuked or beaten it is because beating slaves is in itself
(apparently) funny, and much funnier when it is for no good reason. A slave who
makes wisecracks would be unlikely to be punished, because if the punishment is of
any consequence the result would be to shut him up, eliminating the role of clown that
he had been occupying. A slave who is not really involved in the drama and who has
not done or said anything (good or bad) is more likely to be beaten, on whatever
pretext can be found or for no reason at all.29
One final possible explanation for why the slave at Peace 8191126 is treated so
laxly is that his appearance coincides with the resumption of action after a choral
sequence in which it is claimed that Aristophanes, in addition to discontinuing other
brands of lowbrow humour, such as ridiculing the poor (
, Pax 740) or depicting a Heracles frustrated in his
, Pax
gluttony (
741), has freed the slave characters whom his rivals allegedly brought on stage for the

28 In other words Trygaeus and the slave are talking about Theoria the character at the same
time, and with the same words, as their actors are talking about the pleasures of the festivals she
represents. The master and slave jointly enjoying the sexual benefits of Theoria could represent
the domestic reunification of the fractured
in the same way that the Athenian and Spartan
simultaneous attraction to (respectively the vagina and anus of ) the mute female nude Diallage
(Reconciliation) represents their reconciliation at Lys. 1148, 578.
29 Cf. Nub. 569; Vesp. 12917, 1307; Av. 1311, 1316, 13239, 13345; Lys. 121523.

DANIEL WALIN

37

sake of abuse (Pax 7437).30 As a consequence of the poets recent self-awareness, we

encounter a slave who is treated by his master in an unusually indulgent manner.31
Trygaeus and his slave continue to play jokes off one another. The slave, who has
been the primary joker since he re-emerged on stage at 868, sets up one for his master
by pointing out (metatheatrically) that Ariphrades (whom Aristophanes attacks on
several occasions as overly enthusiastic about cunnilingus) seems to want to be the
one to receive Theoria.32 The master can then make the conventional joke about
Ariphrades (cf. Eq. 12807; Vesp. 12803). But the slave is back in charge of the
humour again at 8913:
{
{

}
}
(Pax 8913)

{Tr.}
{Sl.}

Look at this oven all your own.

How fine it is!
And heres why its so smoky: thats where
the councils pot-props were before the war.

Trygaeus once again prepares the way for his slave by calling Theorias vagina an
oven,33 adapting his diction better to set up the joke, as he did a few lines before with
The slave gets a laugh by explaining Theorias pubic hair as soot. After
these lines the slave suffers a temporarily diminished role; Trygaeus monopolizes the
humorous discourse, not even needing his slave to assist him, for ten lines (894904),
in which he exploits the dual role of Theoria, as woman and symbol of festival, to
produce a number of double entendres. He describes either athletics in extremely
sexual language or (what comes to the same thing) sex with metaphors from athletics.34 Trygaeus then converses with the chorus, and the slave has no speaking role until
922.35
30
can mean simply to put an end to (LSJ s.v. I.2) and may well have that meaning
here (so Sommerstein and Henderson in their translations got rid of and cashier respectively
and certainly to use the term freed in a translation could conjure up for the modern reader
thoughts of abolitionism that we have no reason to suspect in Aristophanes). But it can also
mean to free (apparently often in a metaphorical sense, LSJ s.v. II), and it is interesting to think
about the possibility of that sense here.
31 This explanation should not be pushed very far, for Aristophanes claims not to resort to the
crude methods of his rivals are sometimes tongue-in-cheek.
32 For Ariphrades, cf. Olson (1998), 239 or Sommerstein (1985), 175.
33 Oven is the translation of Sommerstein. Henderson translates
as cooker, while
Olson (1998), 240 translates bake-house. There appears to be some disagreement over whether
this is the place or means (or both) of roasting (
)
34 An interesting discussion of this passage is to be found in F. Garca Romero,
: les mtaphores rotico-sportives dans les comdies dAristophane, Nikephoros 8
(1995), 5776, at 6776.
35 It has been suggested to me that the brief silence of the slave at 894921, after so many lines
of verbal banter between slave and master, may be imputed to the fact that here we have interaction between Trygaeus as citizen and the
, and that the slave as a non-citizen is therefore
excluded. But this neglects the slaves conspicuous presence earlier in the
scene (87794)
and the fact that the text becomes (if possible) even more graphically sexual and comic, not more
serious, after the slave becomes silent. We should not let our understanding of the limitations of
slaves in real Athenian society affect our conception of the roles of comic slaves (not the comic
slave, for there is more than one distinct type) too much. I am more inclined to attribute to
Trygaeus a kind of comic
: he emits a stream of jokes so amusing and so tightly bound

38

AN ARISTOPHANIC SLAVE: PEACE 8191126

The slave shows a certain leadership by beginning the discourse about what should
, 922).36 His phrasing of this line
be done next (
(what should we two do) emphasizes how he and his master are acting as partners.
He does not ask what are my orders or what would you have me do next. When
Trygaeus suggests that the goddess be established with pots, someone (whether the
slave or the chorus is disputed) takes issue with this idea, implying that the goddess
Peace deserves a better offering than one would give to some grumbling (or
contemptible) little herm.37 Trygaeus suggestion that they use a fatted ox (
, 925) is rejected with a pun, lest they have to
, or go to the aid of anyone.
Likewise his idea that they use a pig fat and large is turned down by way of a
passing reference to the piggishness of someone called Theagenes.38 A frustrated
Trygaeus finally asks his interlocutor what they should sacrifice, and it is answered
the Ionic form of the Attic dative
, meaning sheep so that when someone
advocates war in the
, those present will say
in the Ionic fashion.
is said for , or alas (
Apparently this is a pun in which
, 933). There is more humour in that the Ionians had a reputation for
being somewhat soft; by causing the assembly to speak in Ionic fashion, the speaker
imagines that they will act like Ionians as well, being mild in everything else, as lambs
toward each other, and much gentler toward their allies (9346).
We must now ask ourselves who Trygaeus interlocutor is in this passage. RVPCH
assign 922 to the slave, while L and the Aldine editio princeps attribute it to the
chorus.39 The labelling of the speaker in the MSS should be treated as of no more
value than the opinion of a modern scholar, for it is probably derived from the text
itself: it is extremely unlikely that such identifications go back to the late fifth
in 922 makes it clear that the speaker must be the slave: it
century.40 But the use of
seems incredible that the coryphaeus would distinguish himself from the chorus so
completely. Trygaeus does, however, use a plural verb of his interlocutor at 925
), which would appear to be evidence that he is speaking to a group of
(
people (though an unmetrical
is read in RV).41 Olson uses this reading as
evidence for a change of speaker from the slave in 922 to the chorus in 924. In Olsons
text the chorus is Trygaeus interlocutor from this point until the sacrifice begins at
956, despite the fact that Trygaeus uses a singular pronoun ( , 929) a few lines
later.42 Sommerstein, who keeps the slave as Trygaeus interlocutor, explains
together that his slave is awed into silence. After his monologue (894908) ends, the slave remains
in awe for the duration of his masters brief dialogue with the chorus (90921). This fits nicely
with my argument below that Trygaeus forceful appropriation of his name and demotic at 919
(
) shows that he feels threatened by the competition from his slave.
36 Though the attribution of the lines following this one is disputed, 922 itself is safely given to
the slave because of its use of the dual form
Cf. Olson (1998), 245. One recalls this line a
little later, when Trygaeus too strikes a resumptive note with
at 956.
37 For the interpretation of
here, cf. Olson (1998), 245.
38 For a discussion of this person, cf. Olson (1998), 246.
39 Olson (1998), 245.
40 J.C.B. Lowe, The manuscript evidence for changes of speaker in Aristophanes, BICS 9
(1962), 2742 examines the evidence and reaches the conclusion that while it is possible that
changes of speaker could have been indicated by a system resembling that of the parabolus and
double dot, there were probably very few or no identifications of speakers.
41
cannot be kept because it is unmetrical. But it is the reading of our oldest MSS,
so that
may be nothing more than a correction of the received text, in which case there
would be a slim possibility that something different from both readings was original.
42 Olson (1998), 245 does take note of the
and for this reason qualifies his attribution as to
either the chorus or the coryphaeus.

DANIEL WALIN

39

by saying that Trygaeus speaks 924 to both the slave and the chorus.43 He could also
address the audience, or at least the officials in the front row, who, it could be assumed
for the sake of a joke, would be eager at the prospect of sharing the feast that might be
expected to follow the sacrifice of an ox. Either of these explanations of
seems preferable to a situation in which the slave, who has already established himself
as quite an eager clown, initiates a discussion only to fall silent immediately, allowing
the chorus and/or coryphaeus to exploit its comic possibilities.
Olsons other justification for attributing the dialogue to the chorus or coryphaeus
is that the chorus, unlike the slave, can reasonably speak about what goes on in the
Assembly (9313) and what ought to be our attitude toward the allies (9356).44
This last consideration, I think, is of little value as evidence, for comedy is not
reasonable, and a slave who cuckolded the
for sure (874) and possibly his
master (855), and who has felt comfortable making all manner of jokes, obscene and
otherwise, and occasionally showing up his master on stage, certainly seems bold
enough to speak about politics in an offhand way.45 It is undisputed that it is this slave
who saves the sheep for the chorgos by claiming that the altar of Peace is bloodless
(101822). In the process he shows himself to be more cognizant of the proper
religious usages than his master (though piety on stage, as is made clear to everyone in
the audience, is only a pretext for the real-world thrift of saving a sheep). If he has
better judgement about religious matters there and is not afraid to express it, can he
not also speak about politics? Lysistrata speaks about politics at length, as do the
women of Ecclesiazusae, and as women they were no less disenfranchised than a slave.
If we follow Platnauer, Sommerstein and Henderson in attributing the dialogue at
92236 to Trygaeus and the slave, we have a passage in which the slave is in control of
the dialogue. He initiates the conversation at 922, rejects three of his masters
proposals twice with a joke and finally makes his own suggestion, which is also the
, 934). The
grounds for several jokes. He persuades his master easily (
passage would therefore add to the slaves dramatic dominance; in it he is in control of
the humour and the plot development (insofar as there is plot development in
deciding what sort of animal will be sacrificed).
But even if this passage is assigned to the chorus, the characteristics that the slave
displays here may be found elsewhere. For instance, the slave shows the same sort of
leadership and independence from his master at 101722, discussed in passing above,
where he rejects his masters (rather polite) order to slaughter the sheep (
/
, 101718), saying that it is not right (
) and
explaining, when asked, that Peace does not delight in slaughters, for hers is a
bloodless altar (101920). Trygaeus accepts this advice from his slave and acts on it.
Something similar happens at Peace 11224, where the slave responds to Trygaeus
order that he strike the oracle-monger Hierocles (
, 1121) by saying you do it (
, 1122). The slave then explains that
his master should do the striking because he wants to do something else, namely to
).
strip him of his (dishonestly earned) sheepskins (
The slave also corrects his master at 1050. They have just noticed the approach of
the oracle-monger Hierocles, and they are guessing why he is coming and what he
will say. Trygaeus supposes that he has come to oppose the peace (
43

Sommerstein (1985), 177.

Olson (1998), 245.
45 At Ran. 738813, Xanthias and an
and the Athenian citizens.
44

of the house of Pluto chat about the tragedians

40

AN ARISTOPHANIC SLAVE: PEACE 8191126

/
, 10489), but his slave disagrees, claiming
that he has been attracted by the smell of the sacrifice (
, 1050). Both turn out to be right, though the slave more so, for he has
rightly guessed the oracle-mongers reason for coming, while Trygaeus has guessed
about something that will only happen once they have revealed to whom they are
sacrificing (and after they have resisted, for some time, giving him any portion of the
offering).
Even if we were to give no role at all to the slave at 92236, his role in what follows
that scene and its choral interlude is substantial:
{
{
{
{

}
}
}
}

{
{

}
}

{
{
{
{

}
}
}
}

{Tr.}
{Sl.}
{Tr.}
{Sl.}
{Tr.}
{Sl.}
{Tr.}
{Sl.}
{Tr.}
{Sl.}
{Tr.}

(Pax 96273)
And to the spectators throw some of the barley seed.
There.
You already gave them out?
Yes, by Hermes; there isnt one person who has no seed.46
The women didnt get any.
But tonight theyll get some from their husbands. Now let us pray.
Whos here? Where are the many and good?
Let me give it to these guys, for theyre many and good.
You thought them good?
How not? When I poured out so much water, they went to this same place
and stood there. But lets pray ASAP.
Yes, lets pray.

The slave is comically quick in complying with the order (cf. 958, 1042). Continuing
(a grain
his long tradition of sexual humour, he makes a joke out of the word
of barley), which is a slang term for the penis.47 By Olsons division of the lines, it is
the slave who urges that they pray at both 967 and 973.48 If we accept this attribution,
the slave is once more trying to take control of the action, perhaps eager to get to the
prayer, in which he may (depending on whether one believes the attribution of the
MSS) have a joke up his sleeve. Another joke is apparently based on a ritual formula,
, those present were to respond
by which, when the sacrificer asked

46 For the translation seed, which gives some sense of double entendre to the English, I am
indebted to the anonymous reader.
47 Olson (1998), 254.
R explains that
48

Henderson, Platnauer, and Sommerstein give both of these urgings to Trygaeus.

41

DANIEL WALIN

.49 The chorus are many and good, according to the slave, because they have
allowed themselves to be sprinkled with water and did not run away. We should note
that the slave has made two jokes, one of which is about the audience and therefore
has a metatheatrical element.
Trygaeus begins the prayer in a suitably reverent tone, expending three lines on
titles of the goddess and finally asking her to accept the sacrifice. At this point
someone either the chorus or the slave interrupts his prayer:

50

(Pax 97886)

Yes, accept it, much-honoured goddess, by Zeus, and dont do what women in the act of adultery
do. For they too open the door a little and peer out, and if someone notices them, they retreat;
then, if he goes away, they do it again. Dont do any of these things to us anymore.

If the slave speaks these lines as he does in the MSS he is once again introducing a
scandalous joke, for he compares the goddess to an adulterous woman. I think that
this passage is fitting from a character who, as we have seen from many passages, is
preoccupied with the sexual. But Olson (following Brunck) assigns the lines to the
chorus or coryphaeus. He gives two reasons, both of which have to do with what is
appropriate for a slave to do. He quotes van Leeuwens assertion that it is not a slaves
place to pray for the safety of the city (non est serui pro ciuitatis salute precari) and
argues that a description of the sexual pursuit of a free married woman is more
acceptable coming from [the chorus] than from a servile character.51 Neither of these
arguments can hold much water in my opinion. Despite the desire of everyone,
including myself, to establish firm conventions what can and cannot happen in
comedy eleven plays are not really enough to do it. We do not know the limits of
what a slave could do. What if Frogs had not survived? Would any scholar have
imagined a Xanthias? The paucity of our evidence deprives the non est serui argument
of its strength. Rather than asking whether a slave can do x, we should ask whether
this slave might do x. And I think that nothing in this passage crosses some line that
the slave of Peace 8191126 had not already crossed. It is fitting that the slave who
called Theoria and Opora whores (8489) would cast their mistress in the role of an
at 8734, it is
adulterous woman. If he does cuckold his master at 855 and the
in his character to portray the goddess as yet another married woman whom he wants.
It is interesting to note Trygaeus response to this interruption in his prayer.
Despite his reverent tone before the interruption, when he resumes he takes up the
metaphor begun by the slave (or chorus). The goddess is not urged, as one thinks at
first, to be unlike an adulterous woman by being a chaste woman; rather, she is urged
to commit adultery brazenly and without the (affected or real) modesty of an

49

Cf. RV.
Because who speaks these lines is at issue here, I have not labelled the speaker, though I am
using Olsons text and Olson attributes this passage to the chorus.
51 Olson (1998), 256.
50

42

AN ARISTOPHANIC SLAVE: PEACE 8191126

ordinary adulteress. She should quit trying to be secret, just open the door already,
and let her lovers have her. This is the implication of what Trygaeus says at 98790 (
/
/
/
). Once again the master and slave are a team in
wooing, as was implied at 8734. This time, the
probably is meant to include the
whole audience (since presumably everyone wants peace), so that all the men in the
theatre that day were, in a metaphorical sort of way, invited to share a mistress.
In his dissertation on slaves in Aristophanic comedy, Stefanis observes:

52

In Peace the slave, who cooperates with Trygaeus for quite a while, prefigures Xanthias and
Carion both in the extent of his role and in his behaviour.

He notes that the slave occasionally opposes his master and acts as if he were his equal
both of which are traits of Xanthias and Carion and cites some specific examples
) at lines 958 and
in which he speaks back to his master lightly (
1041.53 The first, while it may seem mildly impertinent (
, 958), is nothing compared with other actions and speeches of this slave,
and it is strange that Stefanis chooses it as his first example. The second is not much
different:
{

{
{

}
}

{Sl.}
{Tr.}
{Sl.}

(Pax 103942)
Thats done. Pick up and arrange the thighs. Ill go after the guts and
cakes.
Ill do it. But you should be back already!
Look, Im here. Surely you dont think Im dawdling?

The distribution of these lines is disputed. What I have printed is the distribution of
the MSS, which is advocated by Lowe against the altered arrangements of Beer and
van Leeuwen.54 Arguing that the slave should not give the master an order, Beer
inserts a change of speaker before
, so that the master speaks the second half of
line 1039 and all of 1040. The slave then would speak 1041a (through
), and
the master 1041b. The distribution of van Leeuwen is the same, except that he gives
to Trygaeus as well.
All of this explains why, when one goes to check Stefanis reference to an impertinent remark by the slave (1041), one will find instead a line attributed to Trygaeus, as
52

Stefanis (1980), 151.

To the examples of this slaves impudence cited by Stefanis should be added, I think, Pax
11224, where the slave responds to Trygaeus order that he strike the oracle-monger Hierocles
(
, 1121) by saying you do it (
, 1122). The
slave then explains that his master should do the striking because he wants to do something else,
namely to strip him of his (dishonestly earned) sheepskins (
).
54
J.C.B. Lowe, Some questions of attribution in Aristophanes, Hermes 95 (1967), 5371, at
634. The arguments of Lowe and the reading of the MSS seem to have prevailed. While Coulon,
Cantarella and Platnauer follow van Leeuwen, the more recent editions (Olson, Sommerstein,
Henderson) follow the MSS distribution.
53

DANIEL WALIN

43

in the MSS, in the more recent English editions. Stefanis apparently means to call
talking back lightly the half-line of the slave as in Platnauer (
). I cannot agree that we should necessarily read any level of impudence into such a
remark, but I do agree with Stefanis that if we accept the MSS distribution, there is a
certain feeling that Trygaeus and the slave are equals.55 In this case the slave comes on
stage, reports that the sheep has been slaughtered, and orders his master to take the
thigh bones so that he himself may fetch the rest of the necessary parts of the animal
from offstage. Technically he is giving his master a command, though he is not so
much assuming the role of the master as that of a comrade. One recalls the
partnership of Chremylus and Carion in Wealth.56 The slave is indignant at his
masters fault-finding at 1041 and expresses himself in 1042. This also reminds one of
Carion, who resents the old men who make assumptions about his character
(presumably because of his slave status) at Wealth 2734.
But I find this evidence much less interesting than the constant interjections,
comments and jokes of the slave in previous lines (855, 86870, 8756, 87980, 8913,
all discussed above). For whatever reason, Stefanis does not seem to discuss these.57
There is no substantial comment on this slaves frequent attempts to gain the
sympathy of the audience through initiating, while not being the object of, (often
sexual) humour.58
In an interesting study of names and naming in Aristophanes, Olson observes that
comedy and tragedy are fundamentally different in their presentations of the names
of their characters,59 because tragedy is based on universally recognized myths, so that
to introduce a character by name is to let the audience know everything about him or
her,60 while in comedy a name means little, since the audience must discern the nature
of each character through his or her actions on stage.61 He points out that when one
or any labelling of
watches a comedy on stage, one does not have a list of
characters with a name whenever they speak. Instead, except for certain traditional
figures that would have been recognizable, such as Heracles, Hades or Hermes (and
perhaps, because of his lack of a beard, Cleisthenes), the characters on stage would
55

Stefanis (1980), 151, n. 36.

K.J. Dover, Aristophanic Comedy (Berkeley, 1972), 205 observes that Chremylus and Carion
act more like a pair of friends than slave and master.
57 Stefanis (1980), 115, n. 17 does briefly discuss 8734 (where the slave mentions that he and
his master have screwed Theoria), but he says only that these lines are characterized by
(
) and that they should be attributed to the slave. He
directs the reader to Platnauer (1964), 142, who denies that
can have a sexual meaning.
58 The caveat while not being the object of is quite important, for slaves in Aristophanes often
generate humour at their own expense, so much so that K.J. Dover, Aristophanes Frogs (Oxford,
1993), 434 assigns to slaves in the plays before Ran. only two basic roles (with the exception of
the handling of props): narrating the prologue and eliciting laughter by being hurt, threatened,
or frightened. One can (and Dover does) adduce many examples of this phenomenon in the
earlier plays, such as the slave who is beaten by Philocleon at a party (offstage) and then emerges
on stage to tell the audience about his experience and lament that he does not have the hard shell
of a turtle (Vesp. 12917, 1307), or the silent slaves who are abused for executing orders too slowly
or for being in the way (Av. 1311, 1316, 13239, 13345 and Lys. 121523 respectively). But the
slave at Pax 8191126 does not fit into any of Dovers pre-Xanthian categories.
59 S.D. Olson, Names and naming in Aristophanic comedy, CQ 42 (1992), 30419, at 3046.
60 Cf. Antiphanes fr. 191.58 K:
/
/
/
61 Cf. Antiphanes fr. 191.1820 K:
/
/
/
56

44

AN ARISTOPHANIC SLAVE: PEACE 8191126

have been differentiated only by their dress until they happened to be named (which
did not always happen).
Olsons study of slaves in particular yields results interesting for our purposes. He
comes to the conclusion that, generally speaking, mute slaves are more likely to be
named than those that have substantial speaking parts. He argues that speaking slaves
are denied names in order to prevent them from wielding their names in potentially
subversive ways, while those few substantial actors and speakers who happen to be
given names (for example, Carion in Wealth, Sosias and Xanthias in Wasps) are
always named by their masters in a direct command, and their names do not help
them to dominate the stage.62 The only exception is Xanthias in Frogs. If one accepts
Olsons arguments, it would appear that the slave at Peace 8191126 is not named lest
he be a serious threat to the dramatic dominance of his master during that part of the
play in which they act as partners. It should be noted that Trygaeus name, which has
already been disclosed in his confrontation with Hermes earlier in the play (190), is
, 919).
repeated again quite emphatically in this section (
Does he do this to be sure that no one can mistake who the hero of the play really is,
perhaps including his demotic to assert his citizenship in contrast to the slaves lack
thereof ? If so, this in itself is evidence of how much the slave has been giving Trygaeus
a run for his money in the competition to dominate the stage. That the slave stops
short of actually becoming the hero and must eventually be put down is hardly worth
saying; the same could be said of our most conspicuous example of this kind of slave,
Xanthias in Frogs.63
If Olsons theory is correct, then, this slaves lack of a name indicates that he is a
potential threat or rival. Alternatively, the naming of slaves in comedy is for the most
part mere chance the masters themselves, as Olson recognizes, often go for hundreds
of lines without a name, and this slave (unless he is the same as one of the slaves
earlier in the comedy) has only a few hundred lines to exist. I do not think that a
name would necessarily make him much more powerful, unless he were to use it
himself with a strong sense of self-identity, as Xanthias does with his coinage,
(Frogs 499), which Olson (oddly) does not mention. If he were
named in the course of one of Trygaeus commands, the name would be no more his
than the names of Carion, Sosias and Xanthias (the one in Wasps) are theirs; to be
named in an order is to be given a name, to receive even that from the master.
Xanthias at Frogs 499 and Trygaeus at Peace 919 proudly and with a fully-formed ego
take their names and use them to draw attention to themselves and the identity that
lies behind (better, in) their masks. The slave in Peace certainly does not do this; but
neither does his lack of name preclude him from being a significant forerunner of
Xanthias. Because of the difference between reading a play and seeing it staged in a
time before programmes, his lack of a name is far more likely to prejudice modern
scholarship than to have affected the thinking of the original audience.
CONCLUSIONS
The slave of Peace 8191126, though nameless, proves to be quite a strong character.
He vies with his master for the audiences attention by volunteering all types of jokes,
many of them sexual, one (855) perhaps in an aside. He frequently treats that same
62

Olson (n. 59), 30912.

Xanthias is, of course, the most independent, dominant and interesting slave in
Aristophanic comedy. But even he disappears from the play roughly halfway through at 814.
63

DANIEL WALIN

45

master, and to some extent that master treats him, as an equal and a partner. Often
enough he is temporarily in charge of the action, even if we concede 92236 to
Trygaeus and the chorus. As far as the humour goes, he and his master frequently take
turns setting up one anothers jokes. The slave frankly admires, even fondles, women
; it is evident from the
who are to be betrothed to either his master or the
dialogue that he has slept with the latter, and it is possible that he sleeps with the
former during the play. If the line distribution of the MSS is to be trusted, he also
casts the goddess Peace in the role of his adulterous beloved in a prayer to her. There
are many ways that one could try to explain this slave in terms of the selfconsciousness of the poet after what is said at Peace 7437, the leniency of this
particular master, the exigencies of humour on the comic stage, or the necessity of
reacting out of character to metatheatrical statements. But ultimately, like Xanthias
in Frogs, we are compelled to accept his existence and alter our preconceived notions
accordingly.
University of California, Berkeley

DANIEL WALIN
dwalin@berkeley.edu

WORKS CITED REPEATEDLY

S.D. Olson, Aristophanes Peace (Oxford, 1998).
M. Platnauer, Aristophanes Peace (Oxford, 1964).
A.H. Sommerstein, Peace (Oxford, 1985).
I.E. Stefanis,
    `0 "   %
 (EEThess Suppl. 29, Thessaloniki, 1980).

Classical Quarterly 59.1 4659 (2009) Printed in Great Britain

doi:10.1017/S00098388090000044

46
IN THBROACKES
E CLO U D S
JUSTIN

IN THE CLOUDS: WAS

SOCRATES HIMSELF A DEFENDER OF
SEPARABLE SOUL AND SEPARATE FORMS?

Aristophanes, Clouds 1934

As Strepsiades casts his first uncomprehending glances into the Socratic thinkingshop, he hears that the men bent over are searching the regions below Tartarus. But
what, then, is their arsehole doing, looking to the heavens? Alone by itself it is learn, emphatic in
ing to do astronomy, the Student replies. The phrase
its near-repetition of the ideas alone and by itself, is passed over without comment in
the main editions. It is, however, a phrase that became distinctive of what is usually
thought of as specifically Platonic philosophy used both of the separation of the
from the body, as far as possible, and of the separation of a
soul
Form
from, it seems, things in the everyday world (e.g. Phd. 83b1; cf.
64c58, 65c5d2, 66a13, 79d17). The question arises, therefore, how much these
Platonic ideas may themselves have figured already in the thought of Socrates some
thirty-five years earlier1 despite the standard view, promoted by Aristotle, that they
did not in the slightest.
The phrase
is unremarked in the main editions I have seen of the
play individually,2 as also in the larger editions of Aristophanes;3 lines 1934 are
indeed omitted from some otherwise helpful older editions.4 There are no relevant
scholia5 and there is nothing in Bekkers (anonymously published) compendium of
earlier editions.6 One school edition notes the phrase as philosophical, without saying
in what way;7 probably the most promising note that I have seen is in L.L. Formans

1 The Clouds was first produced at the City Dionysia of 424/3, and the text we have is a partial
revision, from some time between spring 420 and late 417; K.J. Dover (ed.), Aristophanes: The
Clouds (Oxford, 1968), lxxxlxxxi; abridged ed. (Oxford, 1970), xxix. The main revisions that we
know of (from Hypothesis I) do not concern the parts of the play that we will be concerned with.
Scholars mostly date the Phaedo to the years following Platos first visit to Sicily in 388/7, e.g.
W.K.C. Guthrie, A History of Greek Philosophy IV (Cambridge, 1975), 325.
2 K.H. Weise (Leipzig, 1822), I. Bekker (London, 1826), G. Hermann (Leipzig, 1830), F.H.
Bothe (Leipzig, 1830), C.C. Felton (Cambridge, 1841; Boston, 18777 (rev. W.W. Goodwin)), W.S.
Teuffel (Leipzig, 1856, ed. auct. 1863), W.W. Merry (Oxford, 1879), M.W. Humphreys (Boston,
1885), C.E. Graves (Cambridge, 1898), J. van Leeuwen (Leiden, 1898), W.J.M. Starkie (London,
1911), Dover (n. 1), A.H. Sommerstein (Warminster, 1982).
3 F.H. Bothe (Leipzig, 182830, 1845, 4 vols), G. Dindorf (Oxford, 183538, 4 vols in 7);
T. Kock, Ausgewhlte Komdien (Berlin, 185264, 4 vols).
4 For reasons, I imagine, of school decorum: T. Mitchell (London, 1838), with rather full
notes, and H.A. Holden (Cambridge, 18653).
5 Bekker (n. 2), 73131; Dindorf (n. 3), vol. 4.1; or Scholia in Aristophanem fasc. 3.1 (ed.
D. Holwerda, Groningen, 1977).
6 Notae in Aristophanem sedula recensione collatae ex editionibus Brunckii, Reisigii, Beckii,
Dindorfii, Schutzii, Bentleii, Dobreii, Porsoni [] &c. &c. (London, 1829, 3 vols).
7 Rather a philosophic phrase; perhaps intentionally so, though de re ludicra , W.C. Green
(London, 1881).

IN THE CLOUDS

47

edition for H.W. Smyths Greek Series for Colleges and Schools (New York, 1915):
: by itself, independently but it makes no comment on what the

point of the phrase in the context might be.

The need for comment, however, could hardly be clearer. Virtually all the more
respectable English translations get the phrase wrong. Its taking private lessons on
the stars (B.B. Rogers, 18528) suggests almost the opposite of what is meant, in that
the Greek implies a maximum of independence, while private lessons suggests a
relation of scholastic and perhaps financial indebtedness. A succession of translators
have repeated one phrase: It is studying astronomy on its own account (Anon.,
1912);9 Learning astronomy on its own account (Bailey, 1921);10 Its learning
astronomy on its own account (Easterling & Easterling, 1961).11 Unfortunately,
unless the phrase is just another way of talking wrongly, as we have seen of selffinanced tuition, then it is almost entirely obscure what it might mean.12 In lonely
state tis practising astronomy (Starkie, 1911) has a nice tone of mockery, but in
is a phrase used also in a serious
lonely state could never be as
and straight description of the condition in which we are recommended by Plato to
pursue the highest ambitions of philosophy. The best I have found is Hickie (1859): It
is getting taught astronomy alone by itself .13 But while alone by itself is, I think,
exactly right, getting taught makes definite allusion to a teacher, which may rather
may be (and, I think, here is)
defeat the claim to aloneness, whereas
simply is learning or teaching itself.
The French, unfortunately, are no better.14 Bruncks Latin gets the separation, but
not the acting by itself (Quid ergo podex in clum spectat? Seorsum ille astro-

8 The Clouds of Aristophanes: The Greek Text with a Translation into Corresponding Metres
(Oxford, 1852), which appeared anonymously. The translation, with revisions, was later used by
B.B. Rogers in The Comedies of Aristophanes (London, 190216, 6 vols), which keeps the same
phrase, and in his Loeb translation, vol. 1, The Acharnians; The Knights; The Clouds; The Wasps
(London/New York, 1924).
9 The Eleven Comedies, anon. trans. (London, for the Athenian Society, 1912; New York,
Liveright, 1928, 1932) 1.308.
10 C. Bailey, The Clouds Partly in the Original and Partly in Translation (Oxford, 1921).
11 H.J. Easterling and P.E. Easterling, trans. (Cambridge, 1961). So also, with exactly the same
wording here, Sommerstein (1982).
12 The best I can offer would follow the lead of the OED2 (s.v. account 4.): on its own account
would mean for its own interest, at its own risk, for its own sake. But that would surely be a
mistranslation: to be engaged either in astronomy or intellection for ones own interest or ones
own sake is certainly not the same as doing it alone by oneself (which is what the Greek, I think,
means). When Ion talks (Eur. Ion 60810) of how Creusa will now bear her misfortune
, rather than, as before, with the partnership of her husband, he is evidently talking of
her as suffering alone by herself, not on her own account or for her own sake. And if on his
own account carries any suggestion of the pursuit of ulterior personal advantage, then it is
particularly inappropriate for a phrase attached elsewhere (as in the Phd.) to activities that are to
be performed for their own sake and intrinsic worth.
13 W.J. Hickie, trans., The Comedies of Aristophanes: a New and Literal Translation from the
Revised Text of Dindorf (London, 1859). Another exception, with exactly the same wording
(here), is T.J. Arnold, The Clouds of Aristophanes Literally Translated (London, s.d. [1887]), in
the unglamorous series of Kellys Keys to the Classics. R.H. Webb (Charlottesville, VA, 1960)
comes close with Studying on their own, astronomy; as does J. Henderson (Cambridge, MA,
1998) with Learning astronomy on its own.
14 Il apprend de son ct lastronomie, N.L.M. Artaud, Comdies dAristophane (Paris, 18412,
18794). Il apprend pour son compte lastronomie, trans. A.-C. Brotier, corr. L. Humbert (Paris,
1889). Il sinstruit pour son compte dans lastronomie, C. Poyard (Paris, 1898), and with exactly
the same wording, H. Van Daele (Paris [Bud], 1923, 19342). One happy exception, but for the

48

JUSTIN BROACKES

nomiam discit.).15 The early nineteenth-century German versions are more or less
unique in showing a sense of the philosophical use of the phrase elsewhere, but if
anything they exaggerate it. Voss understands the text perfectly well, I think, though
he renders it in free (and today archaic16) terms: Was will der Arsch denn, da zu der
Himmelshh er kuckt? Selbst vor sich selber treibt er da Sternwissenschaft. But
Droysen (Er an und fr sich beschftigt sich mit Astronomie!) and Schnitzer (Das
lernt an sich und fr sich selbst Astronomie) surely go too far.17 The phrase fr sich
is appropriate enough, but an sich is almost meaningless in the context and merely
imports a puzzling aura of the metaphysical in itself . So it seems, for an accurate
rendering we do best with a nineteenth-century crib or school translation; the
German-speakers who hear philosophical echoes also exaggerate them; and in the
absence of any good notes in the main editions, a succession of translators have made
quite inaccurate sense of the passage.
come to mean alone by himself, herself, itself, etc., as I
How does
first.
is of course the
have supposed if indeed it does? Let us take
, the reflexive pronoun, himself, herself, itself.
has
Attic contraction of
the root meaning down, downwards and, like other prepositions, it had in pre-classical
uses an adverbial force.18 An accompanying ablatival genitive was used of the place
from (
, down from Olympus, Il. 1.44), and a proper genitive of the
place to (
, breathing down upon the shoulders [of Asius], Il.
13.385; also down into); an accusative was used of the place through which the action
]
, blood spurted down through his nostrils, Il.
extended ([
16.349). By an extension of meaning (paralleled in the English going down the road)
the last usage became applied also to cases where the action went through something,
, throughout the city, Lyc. 40;
,
but not spatially down (
with the
along the way). With these fundamental uses, the other main uses of

last word: Il sexerce, part soi, la gomtrie, C. Zvort (Paris, s.d. [1889]). (A part soi, seul
avec soi-mme: A. Hatzfeld, A. Darmesteter, Dictionnaire gnral de la langue franaise du
commencement du XVII s. jusqu nos jours (Paris, 18951900), s.v. part.)
15 R.F.P. Brunck, Comdi in latinum sermonem convers (Strasbourg, 1781, 3 vols), repr. in
Aristophanis Comdi Grce et Latine (Paris, 1846). With this phrasing Brunck also loses
any coincidence with the phrasing of the Phd., which in the Latin of, e.g., Ficino runs (at 66a13)
ipsa secundum se ipsam mentis excogitatione sincera utens, ipsum per se quodlibet sincerum
existens studeret venari (in A. Bekker, Platonis Scripta Grce omnia, 11 vols (London,
1826), vol. 10 Platonis dialogi Latine , my emphasis).
16 Aristofanes von Johann Heinrich Voss mit erluternden Anmerkungen von Heinrich Voss
(Brunswick, 1821, 3 vols; 1.211). Vor sich selber, i.e. fr sich selbst. Cf. H. Paul, Deutsches
Wrterbuch (Halle a. S., 1897; Tbingen, 19929), s.v.
vor.
2
17 J.G. Droysen (Berlin, 18358; Leipzig, 1869 ); C.F. Schnitzer (Stuttgart, 1842). And their
successors have as much difficulty as anyone, capturing neither the ordinary meaning nor any
more philosophical associations: Auf eigne Faust betreibt die Himmelskunde Der [d.i. der
Hintere], H. Mller (Leipzig, 1843; 1861); Der treibt Astronomie auf eigne Hand, L. Seeger
(Zrich, 1952); Der lernt auf eigne Rechnung Sternenkunde, O. Seel (Stuttgart, 1963). Unfortunately, auf eigne Faust or Hand means something like on ones own initiative, under ones own
steam, off ones own bat (originally, I presume: by the force of ones own hand or fist), which is not
what the Greek phrase means. For a person, for example, to grieve
(as at Rep.
10.604a3) is for him to grieve alone by himself, not under his own steam; and the error is pretty
much the same when the case is one of learning astronomy.
18 On
: R. Khner, Ausfhrliche Grammatik der griechischen Sprache. 1. Teil: Elementarund Formenlehre. In neuer Bearbeitung besorgt von Dr. Friedrich Blass (Hannover, 1890) 433,
pp. 47580; H.W. Smyth, Greek Grammar (1920; Cambridge, MA, 1956, rev. G.M. Messing)
1690.

IN THE CLOUDS

49

accusative can be traced, I think, as figurative extensions of the idea of what one goes
through, or what one follows the path of. There is a temporal usage (e.g.
, during or through the previous war, Hdt. 1.67); and there are the
various cases that Smyth classes under the headings of Conformity (
, according to the laws, Thuc. 8.2), Ground on which an act is based (
, owing to friendship, Thuc. 1.60), Manner (
, in accord with
, according to nation,
quietness, quietly, Thuc. 6.64), and Distribution (
nation by nation, Thuc. 1.122): all these can be seen, I think, as developments of the
idea of following the path of (the laws, friendship, quietness, the nation), and hence of
what one acts in accordance with, or goes by (as in our going by the book).19
What is it for a person or thing to act
?
of course
means himself, herself, itself, but also by himself, by herself, etc., i.e. alone (LSJ s.v. I.3:
, although alone, Il. 8.99;
, for we are alone, i.e. among
friends, Ar. Ach. 504; usages with which LSJ lists
). When a person
or
, the force of the phrase is,
or the soul acts
then, I think, that it is acting alone by way of itself, alone in accordance with itself
independently following its own path, not using anything else hence, in short, alone
by itself. It is worth noting that itself by itself (as, for example, in Vlastos [n. 21],
25662) is not generally a good equivalent: when Plato talks of a time when the soul
(Phd. 65c57:

may come to be
), he surely means alone by itself , rather than simply itself by itself .
The qualification as far as possible (
, 65c7, repeated 65a1, 67a3, 67c6, cf.
, 65c9) is a sign of how hard the task to be performed is; whereas
the soul is presumably always (if trivially) itself: what is hard is for it to be alone,
approximating the state of death when it will be pure, through separation from the
folly of the body (67a67).20
With the meaning of the phrase clear in general, I shall say only a little about its use
in Plato in particular. The classic text is the Phaedo,21 where
occurs
repeatedly in one form or another, applied to the soul or to Forms as its objects (e.g.
64c7, 65c7, 65d1, 66a1, 66e6, 67c7, 67e8, 70a7, 78d6, 79d1 and 4, 81c1, 83b1 [bis],

19 Cf. M.L. Gill, who in her Parmenides (Indianapolis, 1996) translates

as by itself
and adds a note (to 129a) giving as meaning for that phrase both apart, on its own and in virtue
of, or because of, itself. The latter can be seen as a small extension of the root idea I have
proposed, following the path of itself, a necessary extension indeed when the phrase comes to be
applied not just to the soul but also to Forms.
20 One might try to defend itself by itself, saying that the soul is only fully itself when it is
also in the relevant sense alone, stripped of bodily accretions so as to become what it really, in
its truest nature, is (cf. Rep. 10.611bc). (Thus the soul is, trivially, always [numerically] itself, but
only under conditions of ideal separation from the body does it come to be [qualitatively] itself,
i.e. of its true nature and character.) And there are of course cases where, in English as in Greek,
itself (myself, etc.) has the force of alone. (Thank you, I can do it myself.) But none the less,
it is hard, I think, to hear phrases like using the intellect itself by itself as actually meaning
using the intellect alone by itself (cf. Phd. 66a12); and if one does so, I suspect it is only by
drawing on a special piece of Platonic doctrine, rather than on the common understanding of the
words themselves.
21 D. Gallop translates the phrase alone by itself in his Phaedo (Oxford, 1975; 1993); as does
Burnet, Platos Phaedo, ad 64c6. R. Hackforth (Platos Phaedo, Cambridge, 1955) uses a variety
of phrases: by itself (64c), alone by itself (65c7, 65d1), pure and simple (for
[sc.
], 66a12). G. Vlastos (Socrates: Ironist and Moral Philosopher
[Cambridge, 1991], 25664) has itself by itself, but, though I cannot accept that as a translation,
I think we are largely in agreement on the purposes to which Plato puts the phrase.

50

JUSTIN BROACKES

100b6; also applied to the body: 64c6), with variants (like

, 67d1)
, 82e4;
,
and parallels using other prepositions (e.g.
83a6, cf. 80e5).22 The main discussion is at 64d67b, where Socrates draws out from
Simmias agreement that the body is an impediment to attaining wisdom and that

there are such things as a just itself , a beautiful and a good (

, 65d47) and greatness, health and strength; he argues that a

person will come closest to knowledge of the being of such things who has trained
himself to think most precisely of each object itself, abandoning the senses, as far as
possible, in favour of intellection (65e66a). Using his intellect alone by itself and
), he would undertake the hunt for each reality
unsullied (
[i.e. each Form] alone by itself and unsullied (
) (66a13).23
If knowledge is fully attainable by human beings at all, it can only be after death and
separation from the body (66e67a); and this is the reason why Socrates is unafraid of
the fate ahead of him: those who practise philosophy aright are cultivating dying, and
for them, least of all men, does being dead hold any terror (67e). The same ideas are
invoked later, at 78d80c, in the third main argument for immortality: the soul and
Forms are kin, so, since the Forms themselves are divine, immortal, intelligible,
uniform, indissoluble, it would be proper for the soul too to be indissoluble, or
something close to that (80c). When [the soul] studies alone by itself (
), it departs yonder [into the realm of Forms] towards that which is pure and
always existent and immortal and unvarying, and by virtue of its kinship with it,
enters always into its company, whenever it has come to be alone by itself (
), and whenever it may do so; then it has ceased from its wandering and, when it
is about those objects, it is always constant and unvarying, inasmuch as it is apprehending things of a similar kind [i.e. constant and unvarying]; and this condition of it
is called wisdom, is it not? (79d). That is the condition we may expect death to bring,
and philosophy in this life to approximate. The soul ceases from wandering as it fixes
on unwandering things and the assimilative process at work here turns out in later
writings of Plato to operate in the practice of astronomy itself. Observational astronomy is, in the educational scheme of the Republic, only a pale shadow of the purer
theoretical astronomy studied in problems (Rep. 7.528e530c). But it is itself also,

22

For other uses of the whole phrase, applied to Forms, see e.g. Sym. 211b1 (of the Beautiful
), Parm. 128e6, Tim. 51c1; for the phrase applied to the
soul, see Rep. 6.485d11. For some more ordinary and unmetaphysical uses, see e.g. Rep. 10.604a3
(quoted later in the main text) and Tht. 206a58 (Theaetetus learning to distinguish the letters
each alone by itself,
). The language of Forms existing each
is, I think, absent from the great analogies of Sun, Line and Cave in the Rep., except,
notably, in application to the sun itself (516b) which is uniquely distant and apart from the other
things of the upper realm, perhaps because, if Plato is talking of the survey of a whole realm or
of Forms (e.g. 508c, 509d, 517b; cf. Phdr. 247c), then he may be becoming more impressed
with their relations to each other, and to the Good, rather than their independence. Indeed it is
the interconnection among Forms (rather than their aloneness) that is stressed in the Line
passage, as reason ascends to the Good and then descends again to an endpoint, making use of
Forms alone by way of Forms and onward to Forms (

, Rep. 6.511c; cf. 510b8), where the old incantatory phrasing is adapted now to
express connection rather than separation. Forms as a group are alone or apart (sc. from
sensibles), but they are not individually alone or apart (sc. from each other). For later discussion
of interrelations, see Soph. 255c258c and, for an application of
none the less,
255c1415.
23 Here and elsewhere I draw on the versions of Gallop (n. 21) and Hackforth (n. 21), but with
modifications and variations of my own.

IN THE CLOUDS

51

the Timaeus will say, a direct way to calm the wandering soul: the revolutions of
reason in the stars, when seen by us, will stabilize the variable revolutions within us
(Tim. 47bc; cf. 90cd), as the soul becomes like the unwandering things that it attends
to.
Socrates case for immortality approaches its highpoint (after which a long silence
ensues, Phd. 84c) with an explosion of recommendations for the soul to act
,
,
and
, and to regard as true
. When bound to the body, the soul
only what is, as are the Forms,
, 83a3), as if through
investigates through the eyes and other senses (
a prison, rather than alone through itself (
, 82e4): it is rolling about
in total ignorance (
, 82e5). Philosophy attempts to
release the soul, by persuading it to withdraw from body and senses,
by urging it to collect and gather itself to itself (

, cf. 80e5) and itself to trust

none other but itself (
), whenever, alone by itself (
), it thinks of any reality alone by itself (
); and not to regard as
true what it observes through other things (

), and what varies in various

things; that kind of thing is sensible and seen, whereas the object of its own vision is intelligible
and invisible.
(Phd. 83a6b4)24

The visual metaphor should be noted: the eye of the soul contemplates the Forms.
The same phrase recurs elsewhere with similar force. Perhaps the most interesting
place is Theaetetus 184b187a, where the same opposition is drawn between what the
soul does through other things and what it does through itself and by itself. On the one
hand there is perceiving, which we do with the soul through the senses (184d4); on the
other hand, there is thinking or judging for example, that two colours or sounds are
like or unlike, same or different, beautiful or ugly which the soul does alone through
, 185e1, 6) and alone by itself (

, 186a4;
itself (
187a5). To attain the being of things, and hence have knowledge of them (186ce, that
is, I think, to grasp the nature of, for example, hardness and softness, or of the various
colours and sounds) is the work of the soul acting alone by itself (187a5): something
achieved by reasoning and calculation, and made possible, where it occurs at all, only
,
through a great deal of effort and education (
is used as a phrase to conjure with in the
186c25). Once again,
promotion of higher thought.
of the various figures bent
When the Student comments, then, that the
over is learning
to do astronomy, might his words be a parodic echo
of (an earlier manifestation of ) any or all of these Platonic conceptions of the soul
and its objects?
Of course much of the science and philosophy of the Clouds is absurd, all of it is
comic, and none of it can be taken at face value as a depiction of Socrates own views
or practice. But one of the methods of comedy is to take the language and catch
phrases characteristic of a persons discourse that sound fine and proper in one
context and to set them instead in a place where they are evidently ridiculous. Dover
more vivid;
talks (ad 193) of a resemblance between anus and eye which makes
but it is the inversions as much as the parallels (and indeed the combination of the two
opposite characters) that sustain the comedy here: that it is the arsehole perhaps the

24 Trans. Gallop, with modifications to bring out the Greek phrasing; I substitute true for
real to render
.

52

JUSTIN BROACKES

least upwardly and immaterially ambitious part of the body that is given the task of
learning alone by itself is of course part of the operation of Aristophanes ridicule.
And obviously from comedy built with such traduced materials, it is not going to be
easy to draw out any conclusions about what the original might have been of what we
see only in exaggerated, partly transposed, form.
There is evidently no mention in our passage of a theory of separate Forms or of
any particular view on the nature or separability of the soul. But we have what looks
like a definite reference to a project or practice of withdrawing from the everyday
environment to be intellectually alone and seek knowledge or learning by oneself
and, if that indeed is what it is, then we must ask what such a project might have
amounted to, in whatever form it might have existed in Aristophanes environment. In
Platos Phaedo, of course, the project involved a separable soul attending to separate
Forms, and it was combined with a large-scale opposition between lower things that
are constantly changing and higher things that are not; while the confusion constantly
produced in the soul by lower things was to be alleviated by turning our attention to
higher things, thanks to a process whereby the mind becomes assimilated in certain
respects to the things it attends to. (Those who think on orderly things themselves
become orderly: cf. Phd. 79d, Tim. 47bc, 90cd; cf. Rep. 6.500bd.) But how much less
than all that might the project have involved in the thought of Socrates, or of others in
?
his world, if they used the phrase
We are facing here a version of the traditional Socratic problem: that of trying to
decide how much of what we find in the mouth of Platos Socrates may actually have
been said or thought by the historical Socrates. The internal evidence of the dialogues
is of course hard to disentangle and the external evidence (principally from
Xenophon and Aristophanes) is suspect and tendentious. But views have been taken,
and for serious reasons. At one extreme, A.E. Taylor and John Burnet ascribe to
Socrates all the main views of the Socrates of the Phaedo, including an immortal soul
and the theory of Forms.25 At the opposite extreme, there is, recently, Gregory
Vlastos, and a large tradition developed in part from Hegel, that denies Socrates any
metaphysical views on either subject and insists that Socrates was exclusively a moral
philosopher.26 In the fairly large middle ground stand others, who ascribe some

25 J. Burnet, Platos Phaedo (Oxford, 1911) and A.E. Taylor, Varia Socratica (Oxford, 1911).
Burnet argues: Whatever Plato may or may not have done in other dialogues , I cannot bring
myself to believe that [in the Phd.] he falsified the story of his masters last hours on earth by
using him as a mere mouthpiece for novel doctrines of his own. That would have been an offence
against good taste and an outrage on all natural piety (xixii). Curiously, the theory of Ideas
or Forms that Burnet attributes to Socrates as also to Plato (and credits originally not to either of
them, but to the Pythagoreans) contains no claim of separation of Forms from particulars (xlvi,
n. 2), which Burnet ascribes only to the friends of Forms of Soph. 248a. As will be seen from the
main text, I am not sure that this fits well with the Phaedos talk of a Form as
(e.g. 66a2, 83b1), but that is an issue that I shall not take further here.
26 G.W. von Hegel, Vorlesungen ber die Geschichte der Philosophie, in Werke, hg. v. Eva
Moldenhauer and Karl Markus Michel (Frankfurt a.M., 196971), 18.441515, 520. Socrates
philosophy had an altogether practical aspect: in regard to the personality and method, Platos
picture of Socrates is more accurate, but in regard to the content of his teaching and the point
reached by him in the development of thought, we have in the main to look to Xenophon
(G.W.F. Hegel, Lectures on the History of Philosophy, trans. E.S. Haldane and F.H. Simson
[London, 18926], 1.397, 414). Vlastos (n. 21) argues that the historical Socrates (49) fits ten
main claims, the first two of which are that he is exclusively a moral philosopher (47), and that
he had no such theory as the Platonic theory of separately existing Forms and of a separable
soul which learns by recollecting (48).

IN THE CLOUDS

53

but not all such views to Socrates like Guthrie, for example, who ascribes to him
immortality of the soul but not separability of Forms, arguing that separable Forms
are a Platonic addition that would have seemed to Plato only a trivial inference (a
legitimate projection27) from Socrates interest in definitions, given his faith in the
survival of the soul.
Aristotle has seemed to support a rejection of Taylor and Burnet: it was Plato, he
says, not Socrates, it seems who in his youth became familiar with the Heraclitean
doctrines that all sensible things are ever in a state of flux and that there is no
knowledge about them, and he maintained those views even in his later years.
Socrates, however, was busying himself about ethical matters and neglecting the world of nature
as a whole but seeking the universal in these ethical matters, and fixed thought for the first time
on definitions; Plato accepted his teaching, but held that the problem applied not to any sensible
things but to entities of another kind for this reason, that the common definition could not be
a definition of any sensible thing, as they were always changing. Things of this other sort, then,
he called Ideas, and sensible things, he said, were apart from these, and were called after these
(Met. . 6, 987b19, trans. Ross, rev. Barnes)

Many people accept Aristotles word here. But we should remember that Aristotle was
himself an opponent of Platos separated Forms, and it is not unusual for a person,
when disagreeing, so to speak, with his intellectual father, to claim his grandfather to
be on his side, and the claim may all too easily go unchallenged when the grandfather
died some fifteen years before the speaker and his contemporaries were even born.
The claim that Socrates neglected the science of nature may cite Apology 19bd in
support; pointing in the opposite direction, however, is not only The Clouds but also
the intellectual autobiography of Phaedo 95a99d, in which Socrates declares
himself, when young, to have had a remarkable enthusiasm for the kind of wisdom
) (96a6). Some people, like Vlastos,
known as natural science (
have preferred the Aristotelian view (dismissing the report of Socrates early interest
in natural science as a canard [161], citing Apol. 19bd in support, and passing over
Phd. 96 without any serious discussion at all); while others, like Guthrie, have concluded in favour of the opposite view, that Socrates did indeed have scientific interests
while young.28 This is not the place to attempt to adjudicate that whole dispute. But
what we are currently finding in the Clouds is a virtually unnoticed piece of evidence
of a kind too often ignored:29 it is external evidence (whether ultimately persuasive or

27
W.K.C. Guthrie, Socrates (London, 1971), 33. Similarly E. Zeller, Plato and the Older
Academy (1874), 227: The reality of Ideas seems to him the direct and inevitable consequence of
the Socratic philosophy of Concepts. To derive the existence of Forms simply from a demand for
definitions of terms like pious would surely be, as critics have pointed out, an invalid move; but
it might yet be plausible, and even in part valid, to derive it from the particular kinds of question
that Socrates asked about (e.g.) piety and his particular conception of what the definitional task
amounted to.
28
Guthrie (n. 27), 1005. The status of Phd. 95a99d as biographical evidence is itself not
clear: for a survey of views and a balanced assessment, see Hackforth (n. 21), 12732.
29
Burnet and Taylor have noted other Platonic ideas that are already to be found in the
Socrates of the Clouds: Aristophanes also knows of the spiritual midwifery of Sokrates, for he
has a jest about the miscarriage of a thought. [H]e represents him as a spiritualistic medium,
and he calls the inmates of the Phrontisterion souls, a word which to the ordinary Athenian
would only suggest ghosts. He also ridicules them for going barefoot and unwashed, and speaks
of them as semi-corpses. All that, and more of the same kind, has a sufficient foundation in
what Plato tells us of the Sokratic doctrine of the soul and the practice of death. (J. Burnet,
Greek Philosophy: Thales to Plato (London, 1914) 145.) Burnet is referring, no doubt, to 137

54

JUSTIN BROACKES

not) of a Socratic interest not just in natural science, but also in some kind of special
epistemology of withdrawal from the everyday world in order to acquire learning
about distant things something that (if we do end up attributing it to Socrates)
would certainly go beyond what Aristotle described as a concern simply with ethical
matters, and something that might easily be combined (whether or not it actually was
in the 420s) with metaphysical doctrines on separation of soul and Forms.30 And
Plato at this moment would have been five or six years old, and certainly not the
source.
Might the phrase just be accidental? Might Aristophanes have quite nonchalantly
as Plato himself talks, for example, at Rep.
set a student at a task
10.604a3, of how a man may grieve in the company of his equals or instead alone by
without alluding to anything particularly metaphysical
himself ,
or, indeed, to anything else at all? Quite possibly, if Aristophanes had been talking of

just one student. But there is a whole collection of them (

, 191) searching the subterranean regions, each with an arsehole looking
at the sky; and that a whole collection of such things should be
said, with the innocence of pure literalism, to be acting alone by itself would surely
make little sense. (And it perhaps explains the fact that so many of the translators
have not understood the phrase at all.) If the phrase means anything, it must surely be
an allusion to something that allows it to make better sense than it does literally. The
phrase might perhaps be an echo of something not properly Socratic, or perhaps of a
body of thought other than the one which I have been drawing out of Plato. But with
some searching I have been able to find no other candidate in other philosophers,
, which I shall discuss after briefly considerexcept for some uses of
ing in the next paragraph a quite separate objection, and which actually only adds, I
. The phrase has a
think, to the case for Socratic talk of withdrawal
huge and important role that we have seen in Plato, which, if (perhaps in some lesser
form) it had played also in Socrates, would make perfect sense of Aristophanes lines
as a good piece of mocking parody making fun of a whole body of philosophical
thought about higher thinking and withdrawal from the confusions of the bodily
world, by transposing one of its most spiritually ambitious catch-phrases to one of
the more heavily material parts of the body. And I cannot help thinking that if the
phrase had itself been a catch-phrase of some quite different philosophy, well enough
known to have resonated with Aristophanes audience but only by confusion associable with Socrates, then Plato would hardly have taken it over to play a new and
different, but again resonant, role in his own philosophy, or at least, not without some
comment to distance himself from that rival body of thought. (When the Socrates of
the Apology wishes to distance himself from the impression given by the Clouds, he

(
), with 139; 94 (
);
and 504 (
), with 1024). Taylor (n. 25), in The
, brings
forward a mass of evidence of Platonism in the Socrates of the Clouds, and much of the material
remains impressive, I think, despite some exaggerations, and even after the influential objections
of, e.g., A.M. Adam, Socrates, Quantum mutatus ab illo, CQ 12 (1918), 12139. But in any
case, Taylor makes no mention in his essay of
at 1934 and, though he
mentions the physics of the clouds, connecting them with Diogenes of Apollonia (165 and n.), he
is silent on the main points I suggest below.
30 For a rather different kind of withdrawal, that of the
(Eur. fr. 193, Ar. Eq. 261),
the man who seeks quiet uninvolvement rather than either meddling or fighting in a dirty political
world, see L.B. Carter, The Quiet Athenian (Oxford, 1986), esp. 16273 on Eur.s Antiope.

IN THE CLOUDS

55

does so [19cd], whether entirely fairly or not.) There is no ground for claiming
certainty in the case, but from the new evidence the most plausible conclusion is, I
think, that the phrase is indeed an allusion to a body of Socratic thought, and that
that body of thought is, at least at a moderate level of generality, the kind of thing
that we find later in the Phaedo.
was a phrase that entirely
It would be a mistake to suggest that
by itself needed a special philosophical reading.31 Related phrases using other
prepositions have a perfectly ordinary non-philosophical usage, and our phrase is no
, himself toward himself
exception. In his own death, Ajax acted
(Soph. Ajax 906; cf. OT 1238); Electra talks of being left in the house to lament
, alone and to myself (Soph. Electra 285); we are told that the Barbarian
failed, for the most part,
, himself by reason of himself , i.e. by his
own fault (Thuc. 1.69.5; cf. 6.18.6). And, using precisely our phrase, Ion talks of
, alone
Xuthus wife Creusa bearing her misfortune with bitterness
by herself (Eur. Ion 610), as also Plato talks of a man grieving on his own, in the
passage I mentioned above (Rep. 10.604a3). The point is not that the phrase is in itself
philosophical, but that in our particular context it makes no good sense unless it
alludes to something, so to speak, offstage; and given what the phrase literally means,
along with the remainder of the sentence, how it is later used in Plato, and the fact
that we are supposedly at the gates of Socrates school, the best candidate in the
circumstances is, as it happens, a doctrine of epistemological withdrawal that Aristophanes apparently associated with Socrates. But at this point someone might suggest
a rival view: that parallel idioms like those I have just mentioned show that the
relevant sense is one of separateness and autonomy, rather than of withdrawal or
and the significance of
literal solitude. I am not sure: there is no single force to
these phrases depends of course on the preposition and the case in question. In his
suicide Ajax acts himself toward himself (whether alone or not), and one can talk
indeed of autonomy as salient rather than aloneness; but Electra laments, surely,
alone to herself, and what is in question is aloneness not autonomy. There are places
, himself down upon himself
where a person brings a judgement
(Soph. OT 228; cf. Eur. Heracl. 143), and there indeed one might talk of autonomy.
(Eur. Ion 610),
But when we hear of Creusa bearing her misfortune
the contrast is explicitly with earlier days when in adversity she had had the company
, 6089): the emphasis is
of her husband (
precisely on the aloneness. This last is as good an example as one could hope for of
our phrase in a non-philosophical context at a time close to that of the Clouds, and it
confirms, I think, the general interpretation I offered earlier of
: as
meaning alone by himself, herself, itself. But in that case, I think, aloneness must
indeed be reckoned a standard (though perhaps in special contexts cancellable) part
of the phrases meaning. There are many different ways to act or be alone, but to come
to be alone is, in many central cases, to withdraw, and in our passage in the Clouds the
pretended withdrawal seems to be for the purpose of learning astronomy, an
epistemological task particularly special for Plato. From that point on to say what
kind of epistemological withdrawal might be at issue, for what range of epistemological projects is I think a matter of probabilities and weighing of hypotheses in the
light of larger bodies of evidence; but that some kind of doctrine of epistemological
31 I am very grateful here for the suggestions of an anonymous reader for this Journal, which I
have developed I hope not too freely.

56

JUSTIN BROACKES

withdrawal is being associated with Socrates and held up as a object of fun, seems to
me a hypothesis that sits well with the philological evidence.
It is time to discuss the only other candidate I know that might be cited as a
, alone
possible object of Aristophaness joking the similar phrase
by itself, as it occurs in Hippocratic texts and, more significantly, in Anaxagoras.
or Attic
, is, I think, at least as ordinary a phrase as
(see LSJ, s.v.
, I.1.2). But it occurs in the Hippocratic treatise On Ancient
and
, indeed,32 that led A.E.
Medicine (VM 14, 15) in a context with
and to say it showed that the
Taylor to call the phrase Platos
technical phrases of the Phaedo are not Platos invention but belong to fifth-century
science (Varia Socratica, 215). It would be an exaggeration, however, I think, to
suggest that any very similar doctrine or conception, as opposed to the words
themselves, were to be found in the Hippocratic treatise. The author has talked of
how, for example, the salty, or the acid, may in the human body become separated off
(the same word is used for secretion) alone by itself (
VM 14), thereby causing disease; he doubts, however, that the same thing
happens with the hot or cold, wet or dry (15), or if it does, that it happens in the same
way (1618); he concludes that the principal causes of disease are not the hot, cold,
wet and dry of the new thinkers, but the bitter, salt, sweet, astringent, etc. of the older
physicians. Separation alone by oneself here is no more than occurs all the time with
humours in the body of the sick, and there would not be much promise, I think, in any
suggestion that it was something of that kind that was the object of Aristophanes fun
.
with
, frr. 9,
Anaxagoras, however, talks somewhat similarly of separation (
frr. 6, 8) and of being alone
(fr. 12, three
12, 13, 14, 16, and
occurrences), and with more philosophical significance in a way seems more
promising. In Anaxagoras, All other things have a portion of everything, but Mind
) is infinite and self-ruled, and is mixed with nothing but is all alone by itself
(
(
) (fr. 12). (By contrast, he says that, of everything other
than Mind, nothing can be separated nor come to be by itself (
, fr. 6): they are all, to varying degrees,
intermixed.) And this separation of Mind from non-mental things is declared
essential to Minds mastery (and, in Aristotle, to Minds knowledge): things that were
) would have prevented it (
) from
mixed with it (
controlling anything in the way that it does when it is actually alone by itself (

); it is the finest of all things and the purest (fr. 12; cf. Arist. DA
429a18). It is Anaxagoras whose works we are told Socrates got hold of with such
eagerness and then found disappointing for not giving a larger role indeed to mind
(Phd. 97b98c); and Anaxagoras who is said to have taught Archelaus, who in turn
taught Socrates (DL 2.16). That Socrates should on these topics have modified
Anaxagorean (as well as Pythagorean) materials and put them to further and new use
would not be surprising. And the demand for separation in the Phaedo is made in
rather similar terms, requiring purity (Phd. 67a5b2, c5) and separation (e.g. 67c7, d4)
as a condition for knowledge, and with a psychological theory of the body as an

32

, they have not, I think, discovered any item that is alone

by itself hot or cold or dry or wet, having no share of any other form, VM 15.

IN THE CLOUDS

57

impediment (
, 65a10) when it is grown together or interspersed with the
, 81c6). Of course there would be a modification of doctrine: whereas
soul (
Anaxagoras has Mind (the cosmic mind) alone by itself, it is individual souls that in
Plato are recommended to be alone by themselves; and the object of knowledge is not
bodily things but Forms (cf. 65c9). And there is a significant change in vocabulary,
(or
) to
. (The former is, I think, never used in
from
Plato for the soul setting itself alone by itself , though there are non-philosophically
charged uses at Soph. 217c3 and Prot. 326cd, and it is used briefly of the Form of fire
(51c1,
(Tim. 51b8) before giving way (for Forms in general, in the plural) to
d4); so we have a new technical usage (
,
) that has
more or less completely displaced
.) But then the remarkable thing is
that Aristophanes associates Socrates with the newer phrase: the change, if there has
been one, has taken place in the time of Socrates, not of Plato. And, far from being
just an echo of some general philosophical or scientific usage (of something like
as used in certain medical theories),
in the Clouds would be
an echo of a characteristically Socratic usage connectable with Anaxagoras, perhaps, as also with the Pythagoreans, but only by way of transformation of language as
well as philosophical advance.
Curiously, once we remember that withdrawal in the Phaedo is withdrawal specifically from the confusions of a world of flux and becoming, we can find much in the
same part of the Clouds that may be an echo also of those latter ideas. When the
Clouds finally appear on stage, heralded by Socrates as goddesses who bestow
intelligence and discourse and thought and fantasy and circumlocution (3178),
Strepsiades first response is that he had thought that the clouds were mist and dew
and vapour, not gods; his second response is more comic: if they really are clouds,
what has happened to them, that they look like mortal women? (3401) The answer is
drawn out of Strepsiades himself by what is supposed to be a barrage of Socratic
questioning (345), though (comically) Strepsiades in his innocence quickly turns the
method round and poses questions instead to his questioner (347, 351). The answer is
that the clouds of the sky themselves come to be whatever they want (
, 347), taking the form of an animal corresponding to the
character of people they wish to expose: if they see a shaggy man, they make a
, 350); to expose a thief (
, 351), they
likeness of centaurs (

, 352); when most recently they saw a coward, they

become wolves (
became deer (354); and finally, having seen Cleisthenes, they have become women
(356). There may be a joking transformation here of the kind of idea that shows up in
the Phaedo (also, I think, not entirely seriously), about the reincarnation appropriate
to the various types of soul (81e24): the gluttonous and lecherous will take on the
body of donkeys and similar animals, those who have preferred injustice, tyranny and
, 82a2) will have the bodies of wolves, hawks and kites, while those
thieving (
with everyday but unphilosophical virtue will return to the condition of tame and
social creatures, like bees, wasps and ants (82b58). The transposition in the Clouds is
of course into a different key: for Platos transmigrating souls merely take on the body
of these various animals, whereas the clouds are said literally to become the animals in
question. But that is the real point: these clouds are constantly becoming constantly
changing shape and character just what Plato will emphasize as being the character
of sensible things: continually becoming, not being, they are things that merely
resemble other things.

58

JUSTIN BROACKES

Once there is a chance of taking the clouds as comic echoes of (what we know later
as) the Platonic conception of sensible things, there are other resonances to be heard.
, 375): when
The clouds thunder when rolling about (
filled with water (
, 383) then by necessity they collide and make a noise
because of their density, just as Strepsiades, when filled with soup (

, 386) at the Panathenaea, has suffered disturbances in his stomach

, 386;
, 388).
(
The pattern of ideas of rolling about, when filled with relevant matter, causing
discord and disturbance, reappears in the Phaedo, in the same contexts as we have been
considering (64c67b, 78b82d), to characterize the souls problems when impeded by
the body in the path of true philosophy. To be rolling about, i.e. roaming, by tombs
) is precisely what the
and graves (
soul of the newly deceased does when weighed down by an excess of bodily material
(Phd. 81c11). The soul of each of us indeed is rolling about in utter ignorance
when tied to the body, examining things through the senses (
, 82e433): inquiry through the eyes is full of deceit (

,
82e3). And the effects of constant bodily demands have earlier been described: the
, 66c4) with desires and longings, producing war and
body fills us up (
faction and fighting, clamour and disturbance (
,
66d5; cf. 66a57; 79c7), which rob us of the leisure for the true philosophy that the
(67a1).
soul would pursue
Thus we have two groups of ideas in both our short portions of the Clouds (18499
and 314411) and in some important parts of the Phaedo (64c67b, 78b82d): the
idea of the learner alone by himself considering his distant objects, and, on the other
hand, the idea of things nearby in constant change, rolling about, being filled with
troublesome stuff and producing disturbance. So the question arises again: might the
philosophical conceptions that go with this second cluster of ideas in the Phaedo have
been present in Socrates as well as in Platothe conception (a) of sensible things as
constantly changing, and (b) of such things as an impediment to knowledge? The case
for finding (b) in Socrates may at first seem less strong than with (a). But without (b),
what would be the point of any doctrine of the need for withdrawal
? If it is not that ones everyday environment is in some way confusing and an
impediment (many very different ways, of course, being conceivable), then what
would be the need to withdraw from it? The signs in the clouds of something like (a) in
Socrates can only be taken as suggestive, rather than probative; but if they were
accepted, we would have reason to reject Aristotles suggestion that the Heraclitean
conception of flux left a mark only on Plato and not on Socrates. It is at least clear, I
think, that Aristophanes associated Socrates with a doctrinal catch-phrase that was
crucial to Plato and Platonism some thirty-five years later. To settle exactly which
Platonic ideas might at that earlier time have gone with the phrase and the larger
question, which Platonic views go all the way back to Socrates if a settlement is to

33
is a buzz-word of what is usually counted as Platonic theory of mind and
reality: Rep. 5.479d (

), Tim. 44d, Phaedr. 257a, 275e, Plt. 309a; though the word is also used in
more everyday physics, Tim. 59d.
(Rep. 4.444b, 10.602c, Phd. 66d5 [quoted above]) and
(Rep. 2.381a, Phil. 63d, Phd. 66a57, 79c7) are characteristic terms in Plato, especially
for the confusing effect of the body on the soul, but less exclusively so. The tyrannized soul, like
the tyrannized city, being dragged about by desire, is full of disturbance and regret (
, Rep. 9.577e3).

IN THE CLOUDS

59

be expected at all, given the available evidence would demand a larger investigation
than I can offer here. But the debate is, I think, worth continuing as well as ending
and we do now have more evidence on it than people have noticed.34
Brown University

JUSTIN BROACKES
justin_broackes@brown.edu

34 I am grateful to Mary Louise Gill and Barbara Sattler for conversation and comments, and
to a reader for this Journal for further suggestions.

Classical Quarterly 59.1 6074 (2009) Printed in Great Britain

doi:10.1017/S00098388090000056

60
PLATO ON FORMS AND CONFLICTING
APPEARANCES
SVAVAR HRAFN
SVAVARSSON

PLATO ON FORMS AND

CONFLICTING APPEARANCES:
THE ARGUMENT OF PHAEDO 74A9C6
In Phaedo 73c1 Socrates commences his recollection argument; he had claimed
(72e56) that
(learning happens
to be nothing other than recollection). There are two different sorts of recollection;
things can occasion recollection of similar and dissimilar things (74a23). In the case
of the recollection of similar things, Socrates asks (74a57),
(is it not necessary to experience this in addition: to consider whether
or not this lacks something in similarity to that which is recollected?). The argument
that follows, in 74a9c6, often called the argument from imperfection, is an attempt to
explain what a thing that occasions recollection lacks in comparison with the thing
recollected.
In this argument Plato then introduces Forms. The argument is not intended to
establish that there are Forms; their existence is explicitly assumed. Plato maintains
that we recollect Forms by observing sensible objects. This intellectual process is made
possible by the fact that these two kinds of things are different. So, assuming Forms
exist, they must be different from ordinary objects of experience. The argument is
meant to explain this difference and its significance, namely that we must already have
knowledge of Forms before we perceive ordinary objects, which being different from
them prompt our recollection of them (74e975d2).
But it is controversial in what way Plato takes the two kinds of things to be
different, for a crucial sentence of the argument can be understood in different ways
(grammatically and philosophically). The sentence in question (74b79) is:
(Do not
equal stones and sticks sometimes, being the same, appear equal to one but not to
another?) Does Plato mean that a pair of stones can appear equal to one person and
not to another, or that a stone is evidently equal to one stone and not to another? The
orthodox view has for some time been that the latter option is correct; Plato is not
talking about conflicting appearances. I shall argue for the first option, and in fact
claim that Plato leaves little room for doubt. I shall also suggest what significance
Plato attaches to the difference between Forms and sensibles, and why it makes sense
to take the sentence quoted to refer to the conflict of appearances. The feature that
differentiates Forms from sensibles is what makes Forms knowable in a way that
ordinary objects of experience are not: while a pair of stones may appear equal or
unequal, the equal itself, the Form of equality, invariably appears equal. It is this
invariability that, according to Plato, makes the equal itself knowable as being equal.
Thus, the conclusion of the argument is that, assuming that the equal itself exists, it
can be known that it is equal. Paraphrased the argument looks like this, on my
reading:
(1) Plato assumes that there are two kinds of things that are said to be equal,
ordinary objects of experience like equal sticks and stones and the equal itself.

PLATO ON FORMS AND CONFLICTING APPEARANCES

61

(2) He then claims that we know that the equal itself is equal.
(3) He argues: when we experience sensible equals we come to think of the equal
itself. Since the latter is different from the former, the conditions for recollection
(set in 73c6d1) are met.
(4) Sensible equals and the equal itself are different in that the sensibles variously
appear equal and unequal, while the equal itself invariably appears equal.
(5) It is this property of the equal itself (that of invariably appearing equal) that
distinguishes it from ordinary objects of experience and grounds our knowledge
that the equal itself is equal.
(6) He concludes: If the equal itself exists, it is different from sensibles in that it is
knowable as equal through invariably appearing as equal.
The same claim, I then submit, is stated in the Hippias major.
THE PHAEDO PASSAGE
Phaedo 74a9c6 is probably the first instance of an explicit argument for Forms being
distinct entities. In order to elucidate in which way an item occasioning recollection is
different from the recollected item itself, Socrates asks Simmias (74a912):

We say, I suppose, that something is equal. I dont mean a stick [equal] to a stick or a stone
[equal] to a stone or anything else of that sort, but something different beyond all these things,
the equal itself.

Socrates suggests that the equal itself is distinct from sensible things that are equal.
Further, Plato seems to be saying that the equal itself is equal. The predicate is equal,
then, can apply to a pair of stones and to the equal itself. The distinction between the
non-sensible equal itself and stones is explicable by their being equal in different
ways.1
Although it is generally held that Socrates here distinguishes between sensible
equals and the equal itself, the passage has not been read (as far as I can see) as
claiming that the equal itself is equal. The reason is hardly that such a claim would fit
Socrates badly, for there are enough examples of self-predication in Platos works, and
unmistakably in the Phaedo itself (100c45, cf. 65d46). Further, if one translates the
as We say, I suppose, that something is equal, it is
sentence
quite reasonable to interpret Plato as claiming that the equal itself is equal. There is
no grammatical reason for objecting to the translation above and in fact it is a natural
translation.2
How has the sentence been translated? Assuming that is the grammatical subject
can be understood as predicative or
of the subordinate sentence, the adjective
is copulative: (A) something is equal; such
attributive. In the first case, the verb
1 In a different context (65d4e5) Simmias had already agreed, without argument, that
something is just by itself (65d46), and the implication is that this is justice. In that same
context, Socrates stressed that we do not grasp what is just by itself through the senses but rather
through intellect, a claim he will shortly repeat.
2 R. Loriaux, Le Phdon de Platon. Volume 1 (Namur, 1981 [1969]), 137, comments: Si lon
prend la proposition isolment et si on la traite dun point de vue grammatical, il semble difficile
de traduire autrement, i.e. otherwise than [q]ue quelque chose est gal.

62

S VAVAR HRAFN S VAVARS S ON

is my translation. In the latter case, the verb expresses existence: (B) something equal
.4 Scholars have opted
exists.3 The difference reflects the ambiguity of the verb
5 No grammatical consideration rules out either
for different versions of reading (B).
translation.6 But there is good reason to prefer (A) to (B). Not only does it introduce
self-predication (which to my mind recommends it),7 but Socrates will also immediately and unambiguously state the assumption that the equal itself exists
(74a12b1), i.e. (B):

Should we say that it [the equal itself] is something or nothing? Indeed we should say [that it is
something], said Simmias, by Zeus, most definitely.

Why take such care to make the statement twice? Further, the ensuing explanation of
what Socrates means is puzzling if one opts for reading (B), for when he turns to
sensibles, equal is clearly predicatively used of them.8 And, lastly, Socrates will in
3 Option (B) is taken from C. Rowe, Plato: Phaedo (Cambridge, 1993), 167, 140, who says that
line 65d7 seems to confirm that
and
are to be taken together (ad loc.), namely,
presumably, that the adjective is attributive. I read no confirmation either way from 65d7. He
translates (B) slightly differently, or we say, I suppose, that there is [exists] something equal (ad
loc.).
4 The same options apply to the aforementioned passage of 65d45.
5 Rowe (n. 3) says: there is [exists] something equal. N. White, Forms and sensibles: Phaedo
74BC, Philosophical Topics 15/2 (1987), 197214, says at 197, there is some equal. D. Gallop,
Plato: Phaedo (Oxford, 1975), 119, translates there is something equal (his italics). G.M.A.
Grube, Phaedo, in J. Cooper (ed.), Plato: Complete Works (Indianapolis and Cambridge, 1997),
50100, translates thus. D. Bostock, Platos Phaedo (Oxford, 1986), 66, similarly says there is
something which is equal. This seems to be the more common translation; cf. Loriaux (n. 2), 138,
who mentions other examples. Even if some of the translations mentioned might be taken to
imply self-predication, it is not clear that they do. A related and illuminating passage starts at
103c11, where Socrates asks Cebes:
(You call something hot and
cold?). The tendency has been to translate this sentence as There is something you call hot and
something you call cold (Grube, above), although there is no reason to do so. Rowe (n. 3) translates and explains ad loc., Do you call something hot and cold?, i.e. Is there something you
call hot and [something else you call] cold?
6 Some translators have simply made
, and not , the grammatical subject of the sentence.
Thus
has for example simply been translated there is an equal, i.e. an
equal exists (G.E.L. Owen, Dialectic and eristic in the treatment of the Forms, in G.E.L. Owen
[ed.], Aristotle on Dialectics: The Topics [Oxford, 1968], 10325 at 115, translates thus, italicizing
equal). This translation excludes the possibility of taking equal predicatively. This inclination
is more conspicuous in other translations: equal is something and there is such a thing as
equality. The first translation (his italics) is by T. Penner, The Ascent from Nominalism
(Dordrecht, 1987), 57, and the latter is H.N. Fowlers Loeb version (Cambridge, MA, 1914),
identical to that of R. Hackforth, Platos Phaedo (Cambridge, 1955). These three translations are
not very accurate. They take Plato to be referring, by the word
, to an object, namely the
equal.
7 Understanding the passage as referring to the self-predication of a Form in no way diminishes the desirability of this reading; in fact it should enhance it, because that is what Forms do in
Plato, they self-predicate. This much is even implied in 74d48, which is the conclusion of the
argument. Plato does not seem to take this as a problematic assertion, and it is usually accepted
without argument; cf. M. Frede, Being and becoming in Plato, Oxford Studies in Ancient
Philosophy, supplementary volume (1988), 3752 at 512.
8 It is difficult to see what motivates translating according to (B). Perhaps the reasoning is as
follows: (i) x is equal, but (ii) x is not a sensible equal but rather (iii) the equal itself; hence (iv) x is
the equal itself. But I find this a baffling reading of the passage. The is involved, on this reading,
has to be one of identity; this allows one to take equal in (i) to be the subject of the sentence, and

PLATO ON FORMS AND CONFLICTING APPEARANCES

63

fact proceed to argue for the claim that the equal itself invariably appears equal
(74b7c3), where equal is without doubt predicative of the equal itself. Surely,
then, the distinction he wants to draw here is between two ways these things are equal,
the equal itself and sensibles. It is therefore justified, I submit, to translate as I suggest
above, unless the assertion of the self-predication of the equal itself proves incoherent
within the argument.
Socrates assumes, then, that the equal itself exists. This assumption is so explicit
that Plato deems it fit to add shortly afterwards (76e45) that if these entities do not
exist, then this argument is futile. It does not follow from this claim that the equal
itself is distinct from sensibles, but just that, like sensibles, the equal itself exists, and it
might for all we know be the same as sensibles.9 Plato needs an argument to show up
the difference, and that is what we get.
Having made this assumption, Socrates commences his argument and asks
Simmias whether we actually know what the equal itself is (74b23):
10

Do we also know this [the equal itself], what it is? Certainly, he said.

The phrase
can hardly refer to anything but the equal itself, and
commentators have understood the reference thus: do we know what the equal itself
is? Socrates has already suggested that the equal itself is equal, and got Simmias
agreement. Now he asks whether we actually know this.11 Simmias answers positively.
Certainly, this is a possible reading of the passage. But it might be objected that
Plato does not plainly say this; he does not say that what we know is that the equal
itself is equal; he may have something altogether different in mind. He only asks
whether we know what the equal itself is. Now, the context makes it clear (or so I have
argued) that it is conceded that the equal itself is equal. Socrates has just elicited
Simmias agreement that the equal itself is equal, and immediately embarks on a
discussion of the predicate equal as used of sensibles and Forms (in 74b7c6). So,
given the context, it is perfectly intelligible to interpret the question as I suggest. But
most importantly, there is no other way of making sense of Simmias concession that
we do indeed know what the equal itself is. For what could he be conceding to know
concerning what the equal itself is if not that it is equal? And, anyway, this is precisely
what Plato thinks he knows, that the equal itself is equal. I infer the following claim:
we know that the equal itself is equal. Having elicited the reply that the equal itself is
equal, Socrates immediately asks from where we acquire this knowledge (74b4). He
will elucidate how we do this. Before we turn to Platos elucidation, consider his use of
.
the verb
Plato uses this verb, and then immediately refers to our
in 74b3, and
again in 74c8. He qualifies neither verb nor noun in any way, or gives any indication
that they refer to something different from what he usually calls knowledge. Yet he is
equal in (i) and the equal itself in (iv) to be identical. So: equal = the equal itself and equal
sensible equals. But treating the is as one of identity seems strange, and in fact precluded by
these versions themselves, for they all understand the verb existentially.
9 See White (n. 5), 1989, 211, n. 2.
10 W has this reading, also found in the margins of B and T. It makes no difference to the sense
of the sentence whether we include it or not, but it is a marginal gloss.
11 On the phrase
, see J.D. Denniston, The Greek Particles (Oxford, 19782), 285: Sometimes
, inquires with a certain eagerness: sometimes
means also, and goes close
with an individual word. Cf. e.g. Phd. 94a12.

64

S VAVAR HRAFN S VAVARS S ON

sometimes claimed to be referring not to knowing and knowledge, but something else.
The reason for this claim is that Socrates remarks later that a knowing person would
be able to give an account of what he knows, and not many people can do that
(76b512).12 Hence the knowledge of our passage might be colloquially used in a
humdrum sense; i.e. it is not really knowledge, but perhaps nothing more than
understanding the meaning of the word equal.13
Such evasions are unnecessary at this juncture, and unsanctioned by the text; if
Socrates has in mind by knowledge the understanding of the meaning of the concept
equal, we should infer that knowledge can include such an understanding rather
than infer that Socrates does not have knowledge in mind at all. For the words he uses
and
. Socrates has told us what we know: that the equal itself
are
is equal. Hence one need not expect him to ask what we know the equal itself to be,
since he has already told us; and he does not ask this question.
But one might expect him to ask from where we have acquired the knowledge that
the equal itself is equal. For if he can explain the source of that knowledge, he can
pinpoint the difference between our knowledge of the equal itself and what he had
previously called our knowledge of sensibles (73c68).14 For that is the point of the
argument: to explain the difference (cf. 74a57). In the present context Socrates does
not indicate that we know that equal stones and sticks are equal, but only that we
perceive them thus; the relation between knowledge and sensibles is left unclear.
The next question, then, is directed at the source of our knowledge of the Form
(74b47):

From where have we acquired knowledge of it? Is it not from the things we mentioned just now,
seeing either sticks or stones or some other equal things, we come to think of this [the equal
itself] from them, this [the equal itself] being different from them?

Socrates asks from where comes our knowledge of it, the equal itself. Since he has
just specified the object of our knowledge as not simply the equal itself, but

12 The knowledge Plato discusses in this argument, and has Socrates claim to share with
others, is that, for any F, the F itself is F. If this is indeed knowledge according to Plato one
might ask how can this view be harmonized with what is said in 76b512, that the knowing
person ought to be able to give an account of what he knows? In that passage Socrates is
explaining that knowledge is recollection (in 75b476c13). Before our birth we did have
knowledge of the equal itself. Then Socrates presents two possibilities. At our birth either we did
not forget this knowledge but continued to possess it, or we lost it by forgetting it. If the latter is
the case, we recover this lost knowledge of the equal itself by recollection, prompted by seeing
sensible equals, i.e. we come to know that the equal itself is equal. Then Socrates asks Simmias to
choose between the two possibilities, and helps him by offering an explanation. If the first possibility obtains, we always possessed knowledge, and everyone would then be able to give an
account of what he knows. But this is clearly not the case. Hence we must choose the latter possibility, that knowledge is recollection, prompted by perceiving sensibles. Simmias graciously
exclaims that the only possibility of the first option obtaining, i.e. that knowledge was not lost at
birth, is in the case of Socrates (for a similar explanation, see L. Gerson, Knowing Persons
[Oxford, 2003], 69, n. 17). I see no conflict between our passage and 76b512.
13 As suggested by Bostock (n. 5), 679. The we of 74b2 might then mean people in general
as opposed to the we philosophers, who can give an account.
14 Already in 65d1166a9 Socrates stresses that what each thing essentially is is not the object
of perception but of thought alone.

PLATO ON FORMS AND CONFLICTING APPEARANCES

65

, he must have in mind knowledge of what the equal itself is, i.e. that it is equal.
Again, there appears to be no alternative.
Socrates answers the question from where we get knowledge of the equal itself by
making two claims. First he suggests that, when we see equal sticks and stones we do
in fact come to think of the equal itself (cf. 73c8). Secondly, he claims that the equal
itself, becoming thus the object of our thought, is in fact different from the sensibles,
and presently turns to explain the difference. Thus, the conditions for recollection are
met.
Before we consider Socrates explanation, it needs emphasizing that Socrates is
answering the question From where do we acquire knowledge of it [the equal itself]?
It has been conceded that the equal itself exists, and it has been suggested that we
come to think of it by perceiving sensible equals. But this does not explain what I take
to be the explanandum, i.e. from where we know that the equal itself is equal: It is not
because we come to think of the equal itself that we know what the equal itself is.
Rather, the explanation of our knowledge is afforded by an account of the difference
between the equal itself and sensible equals.
As expected Socrates explains in what way the equal itself and sensible equals are
different, and thereby from where we actually acquire this knowledge of the Form
(74b7c6):

[Socrates:] Or does it [the equal itself] not appear to you different [from sensible equals]?
Consider it also in this way. Do not equal stones and sticks sometimes, being the same, appear
equal to one but not to another? [Simmias:] Certainly. [Socrates:] What then? Is it ever the
case that the equals themselves have appeared to you unequal, or equality inequality?
[Simmias:] Never, Socrates. [Socrates:] Then these things are not the same, he said, these equal
things and the equal itself. [Simmias:] They do not at all appear the same to me, Socrates.

This passage is evidently intended to explain the difference between sensible equals
and the equal itself by claiming that sensible equals variably appear equal and
unequal, but the equal itself invariably appears equal.15
Here, then, we have Plato concerned with conflicting appearances in the process of
introducing Forms. Prima facie Socrates is referring to different perspectives of
15
There are two apparent oddities in this passage. First, Socrates asks whether the equals
themselves (in plural) have ever appeared unequal, and, secondly, whether equality has ever
appeared as inequality. The plural of the equals themselves is probably sufficiently explained as
grammatically following the foregoing plural of equal stones and sticks, as Owen suggests (n. 6),
11415. For the expressions the equal itself, the equals themselves and equality surely all refer
to the same item, as White (n. 5), 2045, maintains, and is generally accepted. For other explanations, see R.S. Bluck, Platos Phaedo (London, 1959), 511, countered by White (n. 5), 214, n. 25,
P. Geach, The Third Man again, Philosophical Review 65 (1956), 728 at 76, reprinted in
R.E. Allen (ed.), Studies in Platos Metaphysics (London, 1965), 26577, and Damascius commentary on the Phaedo (I.302 [Westerink]). The second issue need not be odd either: not only
does equality never appear unequal, but it never appears as inequality either. Plato argues for a
distinction between equality (or the equal itself ) and sensible equals. If they were not distinct, and
equality and sensible equals were identical, then, by the same reasoning, inequality and unequal
sensibles would be identical. And since equal sensibles variously appear equal and unequal (as
has already been granted), equality would variously appear as equality and inequality. For a
discussion of the options advanced, see Gallop (n. 5), 1235.

66

S VAVAR HRAFN S VAVARS S ON

different individuals: sensibles yield conflicting appearances.16 Thus, it happens that x

appears F to one person and not-F to another person. The prevalent interpretation of
the passage rejects this reading. For now I shall pursue my reading without argument.
Now we have additional claims. The Form is different from sensibles in that it never
yields conflicting appearances: The equal itself invariably appears equal and sensible
equals variably appear equal and unequal. Such is the difference between the equal
itself and sensible equals; the equal itself has a certain property that sensible equals
do not have, namely that of invariably appearing equal. This difference is supposed to
explain from where we know the equal itself, i.e. (on my reading) that it is equal. That
such is the explanandum is clear both because Socrates is still answering the question
asked in the previous passage (i.e. from where we know what the equal itself is), and
because he presently says (in 74c79):
(it is definitely from these equal things [sensible equals], being distinct from
the equal [itself], that you have nevertheless derived and grasped the knowledge of it
[the equal itself]?) So: We know that the equal itself is equal because, unlike sensible
equals, it invariably appears equal. In answer to the question asked in 74b4 (
), Plato claims the equal itself invariably appears
equal. Plato has argued that the equal itself is different from sensible equals, explained
in what way they are different, and that we know that the equal itself is equal.
There are two issues that call for an elucidation in this account. First, what sort of a
claim is Plato making when he says that the equal itself never appears unequal, or as
74d4) the equal itself as equal? Is
he later says, that we always experience (
he claiming that it is a necessary truth that the F itself appears as F? In Protagoras
330c2e2 Socrates clearly implies that one cannot think of the F itself as being
anything but F and exclaims (330d7e1):
(Quiet man! How could
anything else be pious if piety itself is not?). Nowhere is there any hesitation in
affirming that the F itself is F. One is tempted to infer that the appearance of the F
itself as F is invariable in that it is a self-evident truth that the F itself is F, a logical
truth, in the Quinean sense, as Benson Mates put it.17 Plato substantiates an
adjective, which he then, as a matter of course, predicates of itself, thus both
objectifying a property and hoisting the property on the object. Ross for one was
unkindly disposed towards this procedure, and stated that the phrase the x-itself
(
) treats the Idea of x as one x among others. The mistake occurs in its
crudest form in Prot. 330c2e2 ,18 the passage we were considering. In the Phaedo
the invariable appearance of the F itself as F is not viewed as problematic.
The other question is this: How should one understand what I have called the
invariability of an appearance? In the Phaedo passage (74c12), Platos words are
fairly general:
(Is it ever the case that the equals themselves have

16

See for instance White (n. 5), 202.

Cf. B. Mates, Identity and predication in Plato, in S. Knuuttila and J. Hintikka (edd.), The
Logic of Being (Dordrecht, 1986), 2947 at 40: for any Greek such a sentence would be a
logical truth, in the Quinean sense that (a) it is true, and (b) every result of substituting another
adjective for its only non-logical constant is equally true. In short, such a sentence would be felt
as obviously and trivially true. Mates goes on to discuss critically the use Plato makes of this
logical truth.
18 D. Ross, Platos Theory of Ideas (Oxford, 1951), 88.
17

PLATO ON FORMS AND CONFLICTING APPEARANCES

67

appeared to you unequal, or equality inequality? Never, Socrates.). It seems to me

that the variability in question may be of various kinds. It may refer to circumstances
in that, for example, a sensible x appears as F compared with y and as not-F
compared with z, or it may appear F at time 1 and as not-F at time 2, while the F itself
is not subject to such variability. Nevertheless, according to the interpretation
adopted above, this passage of the Phaedo places weight on one kind of invariability:
sensible things appear F to one person and not-F to another, while the F itself
appears F to all. Perhaps this is only one way in which the invariable appearance of
the F itself is contrasted with the variable appearance of sensibles; nothing seems to
preclude the possibility of subsuming all cases of such invariability under this
heading. In the Phaedo an object of ordinary experience displays its contrary features
in a situation where it appears to different observers to have a contrary feature. To
compare, consider Platos characterization of the beautiful itself in Symposium
210e212e. There, Diotima offers an account of the ontological status of the
beautiful: it is ungenerated, imperishable and immutable; it is in every way, at all
times, in relation to all things, and everywhere beautiful, which it would not be
(211a47)
(if it
were beautiful for some and ugly for others. Nor will the beautiful appear to him [the
person who gazes at the Form of beauty] as some face or hands or anything that
partakes of a body ).19 Here, sensibles are clearly described as being beautiful
dependently on observers and circumstances.20
I have suggested that Platos argument is intelligible as a reply to the argument from
conflicting appearances. Further, it has the merit of explaining why Plato considered
the claim that the F itself is F so important for the theory of Forms. But, to
generalize, on this reading he holds that, if anything invariably appears F, then we
know that it is F.
This reading has not been adopted by other scholars. In fact, a reading along these
lines is considered by many to saddle Plato with a poor argument, and should
therefore be avoided. The only convincing way of avoiding a reading along the lines
above is to claim that Plato simply is not discussing the conflict of appearances.
Hence, arguments for a different reading are mainly negative. The negative arguments
are the following.

19 The verb appear is here used in its non-veridical sense, since there is no other explanation
of the dative
.
20 N. White, Platos metaphysical epistemology, in R. Kraut (ed.), The Cambridge Companion
to Plato (Cambridge, 1992), 277310 at 28992, discusses the relationship of viewpoint and
circumstances in Platos characterization of the beautiful itself in greater depth. In addition it
should be mentioned that White (n. 5), 200, (above), 2805, and Penner (n. 6) both accept, as I
do, that Plato is discussing the conflicting appearances of sensibles versus the invariable appearances of Forms. White (n. 5), 2078, discusses possible implications of designating the equal itself
differently and addresses them. Penner has termed the inability to confuse the appearances of
Forms and sensibles an incorrigible conceptual state, and the whole argument argument from
incorrigible conceptual states (n. 6), 201. According to Penner, the Forms, in this argument, are
either meanings or things very like meanings (p. 33). As opposed to sensible equals, when
equality appears to us in the appropriate way (i.e. in pure thought), it is utterly stable: it never
appears to be inequality (p. 185); he argues for this interpretation at length at pp. 95121.

68

S VAVAR HRAFN S VAVARS S ON

OBJECTIONS AND REPLY

,
. This might mean, not (i)
In 74b89, Plato writes
appears equal to one [person] but not to another, indicating conflict of appearances,
but (ii) appears equal [relative] to one [thing] but not to another.21 It is alleged that,
if we opt for the first reading, the argument is intolerably bad. In an influential
footnote in his book on Platos Republic, N.R. Murphy may have been the first to
claim that Plato must have meant (ii). For (i) would seem pointless, since we could
infer only that one of the two had made a mistake.22 This observation has been made
many times since.23 Nothing of any epistemological significance follows the observation that appearances conflict, it is claimed.24 Murphys reading of the Greek
phrase is grammatical, but foists a strange view on Plato, as those agree who would
want to read it thus. Consider two points.25
First, on this reading Plato asks whether equal stones and sticks sometimes, being
the same, appear equal to one [thing] but not to another? (74b79). Why would
sensible things only sometimes appear equal or unequal to other sensible things?
Would they not always appear either equal or unequal to other sensible things?26
Secondly, according to Murphys reading the distinction between sensible equals and
the equal itself is that sensible things are equal (or unequal) to some other sensible
thing, while the equal itself is just equal and not equal to anything. In the light of
these objections, Bostock states, It thus seems that [Murphys] interpretation leads
only to nonsense. But for all that, I believe that [Murphys] interpretation is
probably the right one, and Platos doctrine is indeed very peculiar.27 So, were we to
accept Murphys suggestion, we get a peculiar argument instead of a pointless one.
But whatever the merits of the interpretation of Murphy and Bostock, it cannot be
foisted on Plato unless another assumption is made. Here we arrive at a point of

21 There are in fact two other possibilities, usually and fairly dismissed. The first is appears
equal in one respect but not in another, and the latter is appears equal at one time but not at
another. The first of these hardly captures the Greek; cf. Gallop (n. 5), 122, Bostock (n. 5), 74.
The second is supported by a variant reading,

, but this temporal qualification has

already been indicated by
, and is thus redundant, as pointed out by Gallop (n. 5), 122.
22 N.R. Murphy, The Interpretation of Platos Republic (Oxford, 1951), 111, n. 1.
23 It was taken up and developed by G.E.L. Owen, A proof in the Peri Ideon, JHS 1 (1957),
10311, reprinted in R.E. Allen (ed.), Studies in Platos Metaphysics (London, 1965), 293312,
and again by G. Vlastos, Degrees of reality in Plato, in Platonic Studies (Princeton, 1981 [1973]),
5875, first published in R. Bambrough (ed.), New Essays in Plato and Aristotle (New York,
1965), 119. For further discussions sympathetic to this view, see for example Bostock (n. 5),
738, Rowe (n. 3), 169, Loriaux (n. 2), 13943, and A. Nehamas, Plato on the imperfection of
the sensible world, in G. Fine (ed.), Plato 1: Metaphysics and Epistemology (Oxford, 1999),
17191 at 18890, first published in American Philosophical Quarterly 12 (1975), 10517. Cf. also
the detailed discussion of Gallop (n. 5), 1215, and of K.W. Mills, Platos Phaedo 74b7c5,
Phronesis 2 (1957), 12847, and Platos Phaedo 74b7c5, part 2, Phronesis 3 (1958), 4058.
24 So, most clearly, M. Burnyeat, Conflicting appearances, PBA 65 (1979), 69111.
25 Discussed by Gallop (n. 5), 1223, and White (n. 5), 200.
26 One could reply, as David Sedley has pointed out to me, that a sensible thing can be viewed
only in relation to something equal to it, or only in relation to something unequal to it. In that
sense it sometimes appears equal and sometimes unequal, as is perhaps suggested by Phd. 102de
and Tht. 154bc.
27 Bostock (n. 5), 75. Likewise, T. Irwin, The theory of Forms, in G. Fine (ed.), Plato 1:
Metaphysics and Epistemology (Oxford, 1999), 14370 at 153, says that taking the equal itself as
not equal to anything might well appear a nonsensical conception of an equal thing. Owen
defends this line by appeal to Aristotles criticism of Plato. C. Kirwan, Plato and relativity,
Phronesis 19 (1974), 11229 at 11617, is highly critical of it.

PLATO ON FORMS AND CONFLICTING APPEARANCES

69

grammar. In order to make Murphys interpretation work it is crucial that one seize
. If this verb
on another ambiguous Greek phrase, that involving the verb
means appears to be, then Murphys suggestion will not hold. For, first, as Murphys

are
defenders concede, if such is the meaning of the verb, the datives
best taken with the verb.28 Secondly, Murphys suggestion demands that sensible
things be equal to one thing and unequal to another. Plato is not talking about
appearances. His point here is, on this reading, that the equal itself is equal, unlike
sensibles that are equal and unequal; this is the desired distinction. And why would
Plato claim that sensible equals appear to be equal to one thing and unequal to
another thing, if his point really is that they are equal to one thing and unequal to
another? So, if we want to follow Murphy, we have to understand the verb
differently, namely so as to mean, not appear to be, but evidently be. And this is
what most interpreters have done, so that such is now the orthodox interpretation of
the passage.29
The verb in question is ambiguous; if veridical, it is usually followed by a complementary participle, and the meaning is evidently is , while if non-veridical, it is as a
rule followed by an infinitive, meaning appear to be . The problem is that, in the
argument, it is not followed by either.30 We must take the verb to be veridical, Murphy
demands, for otherwise Plato is offering a poor argument. Those who would want to
read the argument in the way presented in the previous section will therefore not only
have to make a case in favour of the non-veridical reading, but also show that the
argument is not pointless, at least not in the alleged way.31 Consider now the
veridical reading, upon which Murphys interpretation rests; we shall see that it is
incorrect because the non-veridical reading is favoured by Plato himself in a summary
explanation of his argument.
First, though, consider Murphys charge that appears to one person [to be] equal
and to another person [to be] unequal only shows that someone is wrong, and not
what Plato would want it to show, namely that there is actually a distinction between
the equal itself and sensible equals, and this distinction is that sensible equals are both

28

Cf. Bostock (n. 5), 734.

It makes for a shorter list to name those who are squarely against this reading, the most
conspicuous of whom are White (n. 5), 200, (n. 20), 2805, and Penner (n. 6). It should be noted
that there are those who accept the translation of the verb as appear to be but still treat the
datives as belonging to equal, so that the translation becomes appear to be equal to one thing
and unequal to another (see e.g. L. Franklin, Recollection and philosophical reflection in Platos
Phaedo, Phronesis 50/4 (2005), 289314 at 304, following A. Silverman, The Dialectic of Essence:
A Study of Platos Metaphysics (Princeton, 2002), 52, 321, n. 8. This way of reading the Greek is
to my mind less appealing than the one adopted in the main text. The dative
close to a
non-veridical
would (I submit) naturally be taken to depend on it. At least one would
need a good reason to interpret the dative otherwise. Nevertheless, even if one were to take the
datives to mean to one thing to another, or even in one respect in another, that would
only change the kind of variability in question, but not that it is a variability of appearances. And
as suggested in the main text (p. 67), Plato may have conceived the variable appearance of
sensibles (and the invariable appearance of Forms) widely enough to encompass different kinds
of appearances.
30 Some have looked to the Republic 47680 in order to find either construction there, to no
avail, except Irwin (n. 27), 154, who does find a veridical use of the verb in Rep. 479b67 and in
Hipp. Ma. 289b57. But the uses of the verb are grammatically ambiguous there in the same way
as in the Phaedo; there is neither an infinitive nor a participle. Irwin does not mention Hipp. Ma.
294c5e4, where the infinitives are clearly to be found, as I shall discuss below.
31 Both Penner and White have done so, although differently from me.
29

70

S VAVAR HRAFN S VAVARS S ON

equal and unequal while the equal itself is just equal. Hence one should translate (ii)
appears equal [relative] to one [thing] but not to another.32
No one doubts that Plato wants to make a distinction between sensible equals and
the equal itself. And a distinction is made on both readings. On my reading the
distinction is that sensibles variably appear equal and unequal, while the equal itself
invariably appears equal. On Murphys reading the distinction is that sensible equals
evidently are both equal and unequal, while the equal itself evidently is equal.33 But
the first reading has the virtue of being an informative answer to the question
from where we acquire knowledge of the equal itself, what it is, i.e. by the equal itself
invariably appearing equal. When sticks and stones appear to us to be equal (or
unequal), we may be in error. When the equal itself appears to us to be equal, we
cannot be in error. That is the important distinction: our cognitive relation to these
appearances is different, how we experience Forms and sensibles. Knowledge is the
issue at this point, not the ontological status of Forms and sensibles, because, as I
claim, Plato is answering the question how we know that the equal itself is equal.
Murphy said that, if we read the passage thus, it is pointless, since we could infer only
that one of the two had made a mistake. But this is all we need to infer.34
But such philosophical reasons for accepting my reading are unnecessary, for it has
gone unacknowledged that, even if an infinitive does not occur in the passage
containing the argument, and thus simply settles the dispute, it actually does occur a
few lines below it, where Socrates summarizes his argument, and thus settles the
dispute. Let us pick up the dialogue where we left off. In 74c7d3, Socrates reiterates
the point that sensibles remind one of the Forms. Then he adds (74d47):

Well then, he said. Do we experience something like this [the following]35 in the case of equal
sticks and the other equal objects we just mentioned? Do they [sensible equals] appear to us to be
equal in the same way as the equal itself,36 or is there some deficiency in their [the sensible
equals] being such as the equal, or is there not?
32

Cf. Bostock (n. 5), 73 and Rowe (n. 3).

It is reasonable to read this view into the argument, because, apart from the fact that other
passages and works indicate that Plato did indeed hold it, Plato implies as much in the present
argument when he states that the equal itself actually is equal in a different way from sensible
equals. But at the same time this reading obscures the immediate context of this particular
argument, i.e. Socrates attempt to answer the question from where we acquire the knowledge that
the equal itself is equal. The reading makes Socrates claim that we know that the equal itself is
equal because we realize that the equal itself evidently is equal while sensible equals are not. This
argument is not interesting, because it does not attempt to answer the question from where we
know that p except by asserting that we know that p because it is true that p. What we would
expect Socrates to offer is a linkage between our knowledge that the equal itself is equal and the
fact that the equal itself is equal. And if he is talking about appearances, we do get this linkage;
the invariability of appearances guarantees knowledge.
34 Further, understanding the verb veridically is semantically strange, for in 74b8 the verb is
apparently contrasted with the phrase
(being the same), so that the intended
contrast would seem to be between how things are and how they appear to be, and not how they
are and how they evidently are. White (n. 5), 201, has more to say on this issue.
35 Socrates is probably referring forwards, as J. Burnett, Platos Phaedo (Oxford, 1911), 57,
thinks and others who express a view on this. But even if Plato is referring backwards, to the
process of recollection, my point is unaffected.
36 The phrase
may be rendered what it is itself or that which is, itself, but in
both cases one is left to supply equal, so that the reference is to the equal itself. Cf. Gallop (n. 5),
229, n. 24.
33

PLATO ON FORMS AND CONFLICTING APPEARANCES

71

Socrates draws together the foregoing argument in order to explain in what way
sensible things are inferior to Forms. The question Socrates has in mind is whether we
experience sensibles as we experience the equal itself: do sensible equals appear to us
invariably to be equal just like the equal itself ? He asks quite clearly, and he uses the
infinitive, whether this is our experience. He does not ask whether sensible equals
evidently are equal just like the equal itself. Nor does he ask whether sensible equals
appear to us to be equal just like the equal itself is equal. This is not surprising, for
Socrates is asking about our experience of these things, not about how things are. The
difference between the equal itself and sensible equals is manifested in the way we
experience them.
Platos use of the infinitive is quite clear; it is clear that he has appearances in mind
in 74b7c6. It might be objected that Socrates is not talking about the same thing here
as in the original statement of the argument. For here he might just be asking whether
there appeared to us to be a distinction between sensible equals and the equal itself.37
But this objection will not do, for he actually spells out what the difference consists in.
What he says is this: sensible equals do not appear to us to be equal in the same way as
the equal itself appears to us to be equal, and in this way our experience of the two
kinds of things is different. Then Socrates finishes his question:

or is there some deficiency in their [the sensible equals] being such as the equal, or is there
not? A considerable deficiency, he [Simmias] said.

It should be clear what the sensible equals lack. They lack invariably appearing to be
equal. Such is the imperfection of the sensible world; it lacks epistemic consistency.38
Plato has set down the necessary and sufficient condition for knowledge of what is
equal: a thing must invariably appear equal. The only thing that fulfils this condition
is a Form, the it itself ; the only thing that invariably appears to be equal is the equal
itself, equality. And this holds good for other themselves, like (75c11d2)

(the beautiful itself, the good

itself, the just, the pious and, as I say, about all those things to which we can attach the
word itself ).39
Sensibles, unlike Forms, suffer conflicting appearances; this is their deficiency.
Since the invariability of Forms appearances grounds our knowledge of Forms, does
not the variability of sensibles appearances imply that we cannot have knowledge of
them, or at least not the same kind of knowledge? When explaining recollection (in
73c4d1), Plato does mention knowledge of sensible things. But having introduced
Forms with the argument of 74a9c6, however, he contrasts the knowledge of Forms,
which we are born with and recollect, with the perception of sensibles, which prompts
the recollection (74e975e7). It is the Forms that are the objects of knowledge, and
37

Such seems to be the understanding of Mills (n. 23), 132.

Plato makes his argument no easier to read when he flanks his question about the difference
between the appearances of Forms and sensibles with the question does it not appear to you
different, and with they do not appear the same to me. Here, undisputedly as far as I can see,
Plato uses the verb appear non-veridically, although there is no infinitive in sight.
39 The invariability of appearances as a condition for knowledge is not confined to Plato. It
seems to be most conspicuously used in the Pyrrhonian tradition; see R. Bett, Pyrrho, His
Antecedents and His Legacy (Oxford,, 2000) and S.H. Svavarsson, Pyrrhos undecidable nature,
Oxford Studies in Ancient Philosophy 27 (2004), 24995.
38

72

S VAVAR HRAFN S VAVARS S ON

sensibles are pointers (75e56):

(would not what we call learning be recovering our own knowledge?). If this is correct, the implication of the argument is: we do not have
knowledge about what sensible things are (equal or unequal), because sensible equals
do not invariably appear equal. This implication of the epistemic status of sensibles,
as opposed to that of the Forms, is only implicit in the Phaedo. It is explicit in the
Hippias major, to which I now turn.40
Socrates and Hippias look for the definition of beauty or the beautiful itself.
Having given up on Hippias suggestions, Socrates suggests that the beautiful itself is
the appropriate. Socrates now asks Hippias (293e11294a2):

See here, then. What do we say about the appropriate: Is it what makes by coming to be present
each thing to which it is present appear beautiful, or be beautiful, or neither?

Hippias gets into trouble. First he chooses the first option: the appropriate is that
which makes things appear beautiful. But Socrates rejects this, since he is looking for
what makes things be beautiful. So Hippias suggests (294c3e4):

But Socrates, the appropriate makes things both be beautiful and appear beautiful, when it is
present. [Socrates:] It is then impossible for things that are really beautiful not to appear to be
beautiful, since what makes them appear so is present? [Hippias:] Impossible. [Socrates:] Then
shall we agree to this, Hippias: that everything that is really beautiful (customs and activities)
both seems and appears to be beautiful always to all? Or just the opposite: that they are unknown,
and there is more strife and contention about them than about anything else, both in private
between individuals and in public between states? [Hippias:] More the latter, Socrates: they are
unknown. [Socrates:] They would not be so, if the appearance of beauty had been added to
them. And that would have been added if the appropriate were beautiful and made things not
only be beautiful but also to appear beautiful. So that the appropriate, if it is that which makes
things be beautiful, would be the beautiful which we are looking for, but would not be that which
makes things appear beautiful. Or, if the appropriate is that which makes things appear
beautiful, it would not be the beautiful for which we are looking. For that makes things be
beautiful, but by itself it could not make things both appear and be beautiful, nor could
anything else.

This passage can be used to elucidate Platos idea of the equal itself in the Phaedo
passage. First, his use of the verb
is non-veridical, as the occurrence of
the complementary infinitive confirms.41 Secondly, people disagree about what is
40
41

Surely an authentic dialogue; see P. Woodruff, Plato: Hippias Major (Oxford, 1982), 93103.
Woodruff (n. 40), 65, and Irwin (n. 27), 154, think otherwise.

73

PLATO ON FORMS AND CONFLICTING APPEARANCES

beautiful, and this is equivalent with things appearing beautiful to some and not to
others. Thirdly, this disagreement shows that it is unknown what things are beautiful.
But it would not be unknown if the things always appeared beautiful. It follows that, if
a thing invariably appears beautiful, it is known that it is beautiful. This is the idea we
found in the Phaedo argument.
Forms and sensibles have a different epistemic status: Plato claims, according to the
above interpretation, that we know the Forms because they invariably appear in the
same way, but we do not know sensibles because they yield conflicting appearances;
Forms are knowable, while sensibles are not knowable. This distinction is epistemic
and not ontological. In light of this interpretation of the Phaedo and Hippias passages I have a suggestion to make regarding a reading of a central passage in the
Republic, where the epistemic difference between Forms and sensibles takes centre
stage. In 476a9480a13 Socrates attempts to persuade the lovers of sights that Forms
exist, having distinguished between these people and true philosophers who do believe
that there are Forms such as the beautiful itself distinct from sensibles. The
sight-lovers, he says, opine and have opinion, while philosophers know and have
knowledge. He then declares his conception of reality (477a34):
(what is completely [F] is
completely knowable [as F] and what is in no way [F] is completely unknowable [as
F]).42 The importance of knowability is complete; what is knowable is coextensive
with what is real.43
Socrates ponders how he could convince the lover of sights, the nominalist, who
does not believe that the beautiful itself is anything, but only that there are many
, here without infinitive
beautiful things (479a5b7). He again uses the verb
or participle, leaving it grammatically unclear whether the verb is used veridically or
not (which he did neither in the Phaedo, the Hippias nor the Symposium):

My dear fellow, well say, of all the many beautiful things, is there one that will not also appear
ugly? Or is there one of those just things that will not also appear unjust? Or is there one of those
pious things that will not also appear impious? [Glaucon:] There isnt one, for it is necessary
that they appear beautiful in a way and also ugly, and the same with the other things you asked
about. [Socrates:] What about the many doubles? Do they appear any the less halves than
doubles? [Glaucon:] Not one.

This
has been taken to be veridical. The reason is the same as in the case of
the Phaedo passage: if it is used non-veridically, nothing can be inferred about the
ontological status of sensibles. Nevertheless a non-veridical reading is grammatical
and, as I will suggest, may be used to make an inference about the ontological status
of sensibles. If read non-veridically, Plato says that the many beautiful things variously appear beautiful and ugly. Let us pursue this reading. In light of this epistemic
42

For the legitimacy of inserting F, see Vlastos (n. 23), 623.

Socrates explains the notion of opinion, how it differs from knowledge and ignorance
(477a678e6). The cognitive faculties that are set over the knowable and unknowable are, respectively, knowledge and ignorance. It is not an innocent move to generalize the connection between
these cognitive faculties and their assumed objects (cf. J. Hintikka, Knowledge and its objects in
Plato, in J.M.E. Moravcsik (ed.), Patterns in Platos Thought (Dordrecht, 1973), 130 at 9.
43

74

S VAVAR HRAFN S VAVARS S ON

feature of the many beautiful things, Socrates asks about their ontological status: is x
then any more F than not-F? (479b910). He gets his answer in epistemic terms
(479b11c5):

They are like the ambiguities one is entertained with at dinner parties for they are ambiguous,
and one cannot understand them as fixedly being [F] or fixedly not being [F], or as both [F and
not-F] or as neither [F nor not-F].

If we adopt the non-veridical reading, Plato claims that, since sensibles variously
appear F and not-F, they cannot be understood as being either F or not-F. We recall the
Republics claim, that the real is knowable and the unreal is not knowable: if
something is not knowable as F, it is not really F. The opinable is not knowable, for it
is
is impossible to understand it as exclusively F or as exclusively not-F. If
read non-veridically, we are afforded a reason why the opinable is not knowable: it
variously appears F and not-F. From our cognitive relationship with objects, on this
reading, Socrates infers a truth about the objects ontological status. The epistemic
inadequacy of the opinable shows its ambivalent ontology. Grammar does not
in this passage. But if Plato uses the
demand a non-veridical reading of
verb non-veridically in the passages of the Phaedo, the Hippias and the Symposium, in
the context of showing the significance of the conflict of appearances, such a reading
of the verb in the Republic passage seems to be reasonable.44
University of Iceland

SVAVAR HRAFN SVAVARSSON

svahra@hi.is

44 My thanks to David Sedley, Eyjlfur Kjalar Emilsson and the journals reader for advice,
elucidations and corrections.

Classical Quarterly 59.1 7590 (2009) Printed in Great Britain

doi:10.1017/S00098388090000068

75
XENOPHON AND THE CHRONOLOGY
OF THE
WAR
JOS
PASCUAL

XENOPHON AND THE CHRONOLOGY OF THE

WAR ON LAND FROM 393 TO 386 B . C .
Although we can glean some isolated facts from other sources,1 we have to rely
primarily on two authors in order to reconstruct the chronology of the Corinthian
War: Diodorus,2 and Xenophon. However, their accounts differ to such an extent that
which one is used as the basis for research determines which of two different chronological timelines is accepted, one that is closer to Xenophon,3 and the other which is
fully supported in Diodorus.4
Using Diodorus as the main guide to chronology presents formidable problems.
For example he dates Agesilaus departure for Asia (14.79.1) as taking place in the
Attic year 396/5, during the archonship of Phormio (14.54.1), when we can be certain
that Agesilaus in fact set sail for Asia in the spring of 396 (Xen. Hell. 4.34, 20), that
is, the Attic year 397/6 (archonship of Suniades). Diodorus also says (14.89.1) that the
trial of the Spartan king Pausanias took place in the Attic year 394/3, when he must
have stood trial immediately after the battle of Haliartus, in the summer of 395 (Attic
year 395/4, cf. Xen. Hell. 3.5.25). In the same way, he reduces the length of time
Agesilaus spent in Asia to two years (a military campaign in 396 and his return in 395)
when the king in fact returned in 394, at the beginning of the third year of his
campaign.5 Similarly (14.83.47, 14.84.285.1), he includes Conons campaign on the

1 For example, Hell.Oxy. 1824 (ed. Chambers); And. 3.1239; Paus. 3.5.110.1; Plut. Lys.
24.2, 29.6.
2 Diodorus account possibly derives from the Hellenica Oxyrhynchia, through an intermediary
author such as Ephorus, cf. H.D. Westlake, Agesilaus in Diodorus, GRBS 27 (1986), 2646; V.J.
Gray, The value of Diodorus Siculus for the years 411386 a.C., Hermes 111 (1987), 73.
3 G. Grote, History of Greece, vol. 9 (London, 1857), 1502; K.J. Beloch, Griechische Geschichte,
vol. 3.1 (Berlin and Leipzig, 1922), 6195.
4 W. Judeich, Die Zeit der Friedensrede des Andokides, Philologus 81 (1926), 14154;
A. Momigliano, Per la storia sulla publizistica della Koin Eirne nel IV secolo a.C., Ann. Pisa 5
(1936), 98103; U. Wilcken, ber Entstehung und Zweck des Knigsfriedens, APAW 15 (1941), 4.
5 According to Diodorus (14.79.13, 14.80.15), Agesilaus reached Asia, led an expedition in
Phrygia from the plain of Cayster to Cyme, and defeated Tissaphernes in Lidia, near Sardis, all in the
same year. However, the two campaigns took place in two different years and Diodorus himself says
(14.79.3) that Agesilaus, between the two, spent the summer in Phrygia and returned to Ephesus in
the autumn. On Agesilaus campaigns in Asia, see C. Dugas, La campagne dAgsilas en Asie
Mineure (395). Xnophon et lAnonyme dOxyrhynchos, BCH 34 (1910), 5895; J.K. Anderson,
The Battle of Sardis in 395 B.C., CSCA 7 (1974), 2753; G.L. Cawkwell, Agesilaus and Sparta,
CQ 26 (1976), 67; V.J. Gray, Two different approaches to the Battle of Sardis in 395 B.C.: Xenophon Hellenica 3.4.2024 and Hellenica Oxyrhynchia 11 (6).46., CSCA 12 (1979), 183200; J.F.
Bommelaer, Lysandre de Sparte. Histoire et traditions (Paris, 1981), 197; P. Cartledge, Agesilaus
and the Crisis of Sparta (Baltimore, 1987), 357; J.G. DeVoto, Agesilaus and Tissaphernes near
Sardis in 395 B.C., Hermes 116 (1988), 4153; C.D. Hamilton, Agesilaus and the Failure of Spartan
Hegemony (Ithaca and London, 1991), 95100; P. Debord, LAsie mineure au IVe sicle. Pouvoirs et
jeux politiques (Paris, 1999), 24350; J. Buckler, Aegean Greece in the Fourth century BC (Leiden and
Boston, 2003), 5869. On his return: P. Funke, Homnoia und Arche. Athen und die griechische
Staatenwelt vom des peloponnesischen Krieges bis zum Knigsfrieden (440/3387/6 v. Chr.)
(Wiesbaden, 1980), 79; Hamilton ibid., 103; R. Seager, The Corinthian War in D.M. Lewis,
J. Boardman, S. Hornblower, M. Ostwald (edd.), The Cambridge Ancient History. Vol. VI: The
Fourth Century B.C.2 (Cambridge, 1994), 101.

76

JOS PASCUAL

Asian coast immediately after Cnidus and the naval expedition the following year to
the Aegean, Laconia and Corinth in the same year, 395/4, in the archonship of
Diophantus (Diod. 14.82.1). In this case, as we shall see, Conon could not have sailed
back along the coast of Asia Minor, then crossed the Aegean, circumnavigated the
Peloponnese, reached Corinth and, finally, anchored at Piraeus in just two months
after August 394, the date of the battle of Cnidus. Furthermore, Diodorus spreads
Thrasybulus naval expedition over two years (14.94.24 and 14.99.45), which is
probably correct, but in 392/1 and 390/89, thus leaving a year between them without
any Athenian naval activity (391/0). He also says (14.97.3) that Diphridas was sent to
Asia before and not after Thibron, when Diphridas was sent specifically to relieve
Thibron, who died in action in Asia in 391 (Xen. Hell. 4.8.19, 212). Particularly in
the case of the middle years of the Corinthian War, Diodorus tends to summarize a
wide variety of different events in a rather haphazard way. Thus he condenses the war
in Corinth between 392 and 390 into practically a single paragraph,6 and does not
consider any of the events of the following two years, which were the archonships of
Antipater (389/8) and Pyrgion (388/7), worth recording.
In short, while recognizing that Diodorus can occasionally supply valuable information for the absolute dating of a particular event, and that he can be used for
obtaining a relative chronology,7 albeit one that is very untidy, it is evident that he has
constructed a chronological timeline that, taken as a whole, is unsustainable and
contains obvious errors.8 For all these reasons, it is hardly a reliable framework on
which to establish the chronology of the Corinthian War.
Xenophons chronology too is far from perfect. In fact, his account contains several
doubtful aspects relating to the campaigns in Corinth in 393 and 392 and the naval
battles the Spartans fought against the Athenians between 391 and 389. However,
despite these drawbacks, Xenophons chronological framework is much more
consistent and reliable. While other sources, particularly the Hellenica Oxyrhynchia,
Andocides, Plutarch and Diodorus can be used for obtaining some additional details,
supplementing Xenophons account is not the same thing as offering an alternative.
There is some consensus with regard to the first two years of the war, 395 and 394.
It is agreed, for example, that the conflict between the Locrians probably the Eastern
Locrians and the Phocians,9 which preceded the outbreak of war, and the battle of
Haliartus took place in 395, in May/June10 and August/September11 respectively, and
6

Diod.14.86.6:

.
G.T. Griffith, The union of Corinth and Argos 392386 B.C., Historia 1 (1950), 245.
8 G.E. Underhill, A Commentary with Introduction and Appendix on the Hellenica of Xenophon
(Oxford, 1900), xlvi; E. Aucello, Ricerche sulla cronologia della Guerra Corinzia, Helikon 4
(1964), 2930; Funke (n. 5), 78.
9 Buckler (n. 5), 76 considers that the Locrians involved were the Eastern Locrians (this is also
my opinion, J. Pascual, Tebas y la Confederacin beocia en el perodo de la Guerra de Corinto
(395386 a.C.), [Diss. Universidad Autnoma de Madrid, 1995], 6849) and that the territory in
dispute was near modern-day Kalapodi.
10 Pausanias (3.9.9) says that the wheat was ripe in Locris at the time of the Phocian invasion. In
Greece wheat can be harvested in June or even in the second half of May (see J. Buckler, The Theban
Hegemony, 371362 B.C. [Cambridge, MA and Boston, 1980], 244), so a date at the end of May for
the war between Locrians and Phocians is possible and a date in June for the Boeotian invasion and
the subsequent Phocian embassy to Sparta is also feasible (Xen. Hell. 3.5.4). These events would have
occurred during Phormios archonship (396/5), which ended in the last days of June 395 (S. Accame,
Ricerche intorno a la Guerra corinzia [Naples, 1951], 467). Diodorus (14.81.13) places all these
events in the archonship of Phormio (396/5), although Haliartus must have fallen in the archonship
of Diophantus (395/4) that followed.
11 The archonship of Diophantus probably lasted from 27 June 395 to 16 July 394. The
7

XENOPHON AND THE CHRONOLOGY OF THE WAR

77

394 was the year of the battles of Nemea (second half of July),12 Cnidus (end of
July/first days of August)13 and Coronea (16/17 August).14 Disagreement mainly
concerns the progress of the fighting on land in Corinth between 393 and 39015.
Thus, since Xenophons account is the basis for determining the chronology of the
Corinthian War, an investigation into the system he used for dating events can help us
establish its chronology accurately, and thus resolve some of the more doubtful and
controversial issues, especially the events relating to the war on land between 393 and
386, which will be the focus of our study.
Firstly, Xenophon chose to give entirely separate accounts of the war on land and
the war at sea. Hence he describes the events that took place on land, without
interruptions, from Hell. 4.4.1 to 4.7.7 and then, from Hell. 4.8.1 to 5.1.35, he tells us
about what happened at sea. Internally both parts follow a chronological order;
however, the basic problem lies in combining, synchronizing and harmonizing the two
parts, between which Xenophon makes little connection, since he uses only very vague
references to relate one to the other16.
Secondly, in general Xenophon appears to have used a year divided into seasons as
the basis of his chronological system, like Thucydides, who probably used a solar

Lacedaemonian embassy to Thebes, the Athenian embassy to Sparta and Lacedaemonian preparations against Boeotia, which would have occupied the month of July 395, probably took place at the
beginning of this archonship (Hell.Oxy. 13.15; Paus. 3.9.911). The alliance between Boeotians
and Athenians must have been signed at the end of July or beginning of August 395 (M.N. Tod, A
Selection of Greek Historical Inscriptions, vol. 2 [Oxford, 1948], no. 101, 1415; P.J. Rhodes, R.
Osborne, Greek Historical Inscriptions 404323 B.C. [Oxford, 2003], no. 6, 3841). The battle of
Haliartus would have been fought in August or during the first half of September 395 (Beloch [n. 3],
6970). The anti-Spartan alliance was formed in the autumn/winter of 395/4 (Diod. 14.82.14) and
the campaign waged by the Argives and Boeotians in central Greece and Thessaly (Diod. 14.82.59)
can be dated to the autumn of 395 or, more probably, the early spring of 394.
12 We have three different inscriptions relating to the battles of Nemea and Coronea. IG 22 5221
is a small fragment of a much larger inscription which listed by tribes those who fell in the two battles,
referred to in the document as Corinth (that is Nemea) and Boeotia (that is Coronea), cf. P. Harding,
Translated Documents of Greece & Rome. 2. From the End of the Peloponnesian War to the Battle
of Ipsus (Cambridge, 1985), no. 19 A, 1920; Rhodes, Osborne (n. 11), no. 7, 423. The fact that
the two battles are listed together could be because they were fought during the same archonship. IG
22 5222 probably commemorates the cavalrymen of a single tribe, the Acamantis, who fell in the
battle, since one of the dead is listed as Dexileus, commemorated in the following inscription. Once
again the dead in both battles are listed in a single inscription (SIG3 131; Tod [n. 11], no. 104, 1820;
Harding, ibid., no. 19 B, 20). IG 22 6217 is the famous funeral inscription of Dexileus, a horseman
who died in the archonship of Euboulides in Corinth (SIG3, 130; Tod [n. 11], no. 109, 201;
Harding, ibid., no. 19 C; Rhodes, Osborne [n. 11], 403). We know that the battle of Coronea,
fought in the middle of August 394, took place in the Attic year 394/3, that is, the archonship of
Euboulides. The archonship of Euboulides lasted from 17 July 394 to 5 July 393, so the battle of
Nemea, before Coronea, could have been fought in the second half of July, in the first few days of this
archonship (cf. Beloch [n. 3], 72).
13 The naval battle of Cnidus would have taken place at the end of July or more probably in the
first few days of August, with sufficient time for Agesilaus to receive the news of the Lacedaemonian
defeat at Cnidus in Boeotia on 14 August 394, the day of the partial eclipse (Xen. Hell. 4.3.10; cf.
Lysias. 29.28; Beloch [n. 3], 70; Hamilton [n. 5], 109).
14 The battle of Coronea (Xen. Hell. 4.3.15; Plut. Ages. 18) took place immediately after the
eclipse, probably on 16 or 17 August 394. Afterwards Agesilaus went to Delphi for the Pythian
Games celebrated at the end of August or beginning of September 394 (Plut. Ages. 19.3), at the same
time as the polemarch Gylis attacked Locris, probably Eastern Locris (Xen. Hell. 4.3.213).
15 The chronology of the war at sea between 391 and 389, which is outside the scope of this study,
also presents considerable problems.
16
Funke (n. 5), 77.

78

JOS PASCUAL

calendar divided into seasons,17 perhaps derived from the Euctemons parapgmata.18
Using this system Xenophon treated the year as beginning in March/April and ending
around February/March the following year19. Although he mentions four seasons, his
basic division was between summer in the loosest sense, that is, the period for
military campaigns, and winter, the season when citizen armies were demobilized. In
the case of Thucydides, summer extended from the appearance of Arcturus (6
March) to that of the Pleiades (8 November) and, in a somewhat similar fashion
Xenophon treated the campaigning season as lasting a maximum of seven or eight
months from the beginning of March to the end of September.20 Hence each year the
armies were mobilized around March/April, or even later, in May/June, and disbanded no later than the end of September.
More serious is, thirdly, his pro-Spartan bias. Xenophon concentrated his account
on the events in which Sparta and his Spartan heroes Agesilaus21 and, to a lesser
extent, Teleutias, were involved, and consequently he preferred to describe the events
that took place on land and in particular those that occurred in the Peloponnese, but
displayed little interest in the war at sea and paid no attention to the way the situation
was developing in Thessaly, central Greece and the Northwest.22
These pro-Lacedaemonian sympathies also have to be taken into account when, in
his account of the war on land, he refers to the following mobilizations and their
corresponding demobilizations from the beginning of the war.
As we can see (cf. below Table 1): Xenophon mentions only mobilizations
and demobilizations involving Lacedaemonian contingents and their allies, and
never those of their adversaries. These mobilizations and demobilizations of the
Lacedaemonian armies constitute, in my opinion, one of the crucial elements for
establishing the chronology of the Corinthian War.23
We have eight mobilizations and eight demobilizations for the ten years between
the outbreak of the Corinthian War and the Kings Peace, that is, between 395 and
386. Unless Xenophon made an improbable series of errors, this means that in one or
more years the Lacedaemonian Army was either not mobilized or did not cross the
borders of the Lacedaemonian state.
M1D1 and M2D2 took place in 395 and 394 respectively. After 394 the problems
of dating become more difficult. Xenophon refers to the demobilizations, D3 and D4
(Hell. 4.4.13 and 4.4.19), that related to two campaigns before the Isthmian Games

17 A.W. Gomme, A Historical Commentary on Thucydides Vol.III. The Ten Years War, Books
IVV 24 (Oxford, 1967), 699715.
18 W.K. Pritchett, B.L. Van den Waerden, Thucydidean time-reckoning and Euktemons seasonal
calendar, BCH 85 (1961), 1752; W.K. Pritchett, Thucydides V 20, Historia 13 (1964), 2136;
R. Hannah, Greek and Roman Calendars. Constructions of Time in the Classical World (London,
2005), 5262.
19 G.E. Underhill, The chronology of the Elean War, CR 7 (1893), 157.
20 See M. Amit, Athens and the Sea. A Study in Athenian Sea-power (Brussels, 1965), 27.
21 Cawkwell (n. 5), 65; J.-C. Riedinger, tude sur les Hellniques. Xnophon et lHistoire (Paris,
1991), 203.
22 His more serious errors include, for example, not mentioning the formation of the antiSpartan coalition in the winter of 395/4 (Diod. 14.82.14) or the Argives and Boeotians
expedition to central Greece and Thessaly (Diod. 14.82.59).
23 In practice, Xenophon also used the Lacedaemonian year which ran from October to the end of
September of the following year and began with the first or second full moon before the autumn
equinox, depending on whether the preceding year consisted of 12 or 13 months (cf. L. Pareti, Studi
minori di Storia Antica, vol. 2 [Rome, 1961], 24, 213, 269).

XENOPHON AND THE CHRONOLOGY OF THE WAR

79

Table 1
Mobilizations (M)

Demobilizations (D)

M1. Hell. 3.5.6. The ephors order mobilization. D1. Hell. 3.5.225. Withdrawal of the
Lacedaemonian army after the battle of Haliartus.
M2. Hell. 4.2.9. The ephors order mobilization. D2. Hell. 4.4.1. Agesilaus disbands the army.
There is a partial eclipse on 14 August this year
(394) and the Pythian Games are held.
D3. Hell. 4.4.13. Praxitas, the Lacedaemonian
harmost in Sicyon, disbands the army.
M3. Hell. 4.4.19. The Lacedaemonians
undertake an expedition against Argos (and
Corinth).
M4. Hell. 4.5.1. The Lacedaemonians once
again undertake an expedition against
Corinthian territory.

D4. Hell. 4.4.19. Agesilaus disbands the army.

D5. Hell. 4.5.18. Agesilaus withdraws to Laconia

with the army. The Isthmian Games are held this
year.

M5. Hell. 4.6.13: Lacedaemonian mobilization

D6. Hell. 4.6.12, 14. When autumn arrives,
against the Acarnanians.
Agesilaus withdraws from Acarnania.
M6. Hell. 4.7.1. Lacedaemonian mobilization
against the Acarnanians, who capitulated
before the attack.
M7. Hell. 4.7.2. Lacedaemonian mobilization
against the Argives.

D7. Hell. 4.7.7. Agesipolis, the Spartan king,

withdraws the army.

M8. Hell. 5.1.33. Agesilaus persuaded the

ephors to order the mobilization.
D8. Hell. 5.1.35. General demobilization of the
army and navy.*
*But only the Lacedaemonian army was actually mobilized (cf. Hell. 5.1.33, 36).

(Hell. 4.5.1). Irrespective of whether they took place in 393, 392 or 391, these demobilizations, together with that of 394 (D2), make it impossible for the Isthmian Games
mentioned by Xenophon (Hell. 4.5.2), which were held biannually, to have been held
in May/June 392. In fact, if the Isthmian Games took place in 392, the Lacedaemonian army would have been disbanded only once, at the end of 393. So we can be sure
that Agesilaus campaign, which was fought while the Isthmian Games were being
held, took place in 390.24 So D5 (Hell. 4.5.1819) must have taken place at the end of
390 and the mobilization that preceded it, M4 (Hell. 4.5.1), also related to that same
campaign in 390.
Between the battle of Coronea (394) and the Isthmian Games (390) we need to
establish the campaigns of 393, 392 and 391 for which, as we have said, we have
references to only two demobilizations (D3 and D4). Since Xenophon refers only to
the Lacedaemonian army, we can advance various hypotheses: either Xenophon has
compressed into two years the events that took place in three, or at least one year saw
24 G.E. Underhill, The Chronology of the Corinthian War, Journal of Philology 22 (1894), 133
dates the Isthmian Games to c. April 390 and Beloch (n. 3), 86 to May/June 390.

80

JOS PASCUAL

no major armed confrontations, or the Lacedaemonian army was not mobilized or

did not fight in the Isthmus. Moreover, together with the reports of two demobilizations in these years, 393 to 391, we know of only one mobilization, M3 (Hell.
4.4.19), which related to Agesilaus expedition against Argos, and its subsequent
demobilization, D4. In short, for three years, 393, 392 and 391, we have one mobilization (M3) and two demobilizations (D3 and D4).
After describing the demobilization of September 394, Xenophon continues with
his account, but where we would expect to find a new mobilization, he says (Hell.
4.4.1):
, After this the Athenians, Boeotians, Argives and their allies now based
themselves on Corinth and carried on the war from there, and the Lacedaemonians
and their allies based themselves on Sicyon. He goes on to describe the massacre of
the Corinthian oligarchs, which took place on the last day of the Eucleiae, that is,
about the month of March. Later some exiles return (Hell. 4.4.5) and the union of
Argos and Corinth, probably a form of sympoliteia, takes place (Hell. 4.4.6).25 After
that the returning Corinthian oligarchs helped Praxitas,26 the Lacedaemonian harmost of Sicyon, enter the city through one of the gates in the Corinthian Long Walls,
which connected Corinth with its port, Lechaeum (Hell. 4.4.78), where Praxitas took
up position. At this point Xenophon (Hell. 4.4.89) tells us about Praxitas contingent, which was made up of a Lacedaemonian mora, the Sicyonians and some
hundred and fifty Corinthian exiles (
). Thus Praxitas has neither the Lacedaemonian army nor the bulk of the allies as a whole under his command. Praxitas wins
the battle of the Corinthian Long Walls that follows, the Lacedaemonians occupy
Lechaeum and the Boeotian contingent garrisoning the port is annihilated (Hell.
4.4.912). Immediately after the battle, Xenophon says (Hell. 4.4.13), consistently,
that the Lacedaemonians allies, and not the Lacedaemonians themselves, came to the
aid of Praxitas. Then they demolished two sections of the Long Walls, west and east
respectively, sufficient to allow an army through, took Sidus and Crommyon, and
proceeded to fortify Epieiceia. Once they had done this, Praxitas disbanded the army
and returned to Lacedaemonia (Hell. 4.4.13).
Obviously Xenophon considers this as marking the end of the year. From his
account it can be deduced that the Lacedaemonian army as a whole did not take part
in the campaign, nor was it mobilized to fight in Corinth, where only the allies and a
Lacedaemonian mora fought. At no point in the campaign does Praxitas appear to
have led a Lacedaemonian army, and the final phase can be interpreted as meaning
that Praxitas disbanded the allied army and not the Lacedaemonian army and that he
himself returned to Lacedaemonia either alone or with his mora, which was thus
relieved.27
If the massacre of the Corinthian oligarchs occurred in March, some time must
have elapsed for the internal situation in Corinth to have stabilized sufficiently for the
exiles to return; the sympoliteia between Argos and Corinth was probably established
25 N. di Gioia, LUnione Argo-Corinto, Contributi dellIstituto di Storia Antica, vol. 2
(Milan, 1974), 39 dates the union to between March and August 393.
26 Cf. P. Poralla, Prosopographie der Lakeidamonier bis auf die Zeit Alexanders des Grossen
(Chicago, 1985), 170: Praxitas is a Lacedaemonian polemarchos, stationed with his mora in Sicyon
in 393.
27 On the relief of the Lacedaemonian morai, see, for example, Plut. Pelop. 16.

XENOPHON AND THE CHRONOLOGY OF THE WAR

81

a little later and Praxitas campaigns fought in the second half of the summer, no
earlier than July/August.
Just as in Hell. 4.4.1, where, after Praxitas demobilization, we would expect to find
the next mobilization, Xenophon (Hell. 4.4.14) claims that

, From this time on neither side sent out

large expeditionary forces. Instead, the cities in each alliance sent contingents for
garrison duty either in Corinth or in Sicyon and these forces merely guarded the
fortifications. Both sides, however, employed forces of mercenaries and used them
vigorously in carrying on the war. He therefore explicitly says that the two sides
armies were not mobilized, at least at the beginning of the year, which clearly means
the Lacedaemonian army was not mobilized and only a mora had remained stationed
at Lechaeum (cf. Hell. 4.4.17). Xenophon then goes on to give an account of
Iphicrates incursions into Phleious and Arcadia with mercenary peltasts, (Hell.
4.4.1516) and a series of raids around Lechaeum and Corinth (Hell. 4.4.17). Finally
) and rebuilt the sections
the Athenians arrived in large numbers (
of the Corinthian Long Walls destroyed by Praxitas (Hell. 4.4.18). Between these two
events, the fighting around Lechaeum and the reconstruction of the Walls, the allies
had reoccupied Corinths port, since this was subsequently retaken by Agesilaus
(Hell. 4.4.19).28
After the reconstruction of the Walls, Xenophon (Hell. 4.4.19) says that the
Lacedaemonians undertook an expedition against the Argives under Agesilaus
command (M3). After laying waste part of the Argolid, Agesilaus went on to Corinth
and recaptured Lechaeum; he attacked by land, his half-brother Teleutias providing
him with naval support. Once Lechaeum had been taken,
, Agesilaus disbanded the army of his allies and led the home army back to Sparta. Hence this was a
full campaign by the Lacedaemonian army, with its respective mobilizations and
demobilizations (M3 and D4). This expedition predates Agesilaus campaign during
which the Isthmian Games of May or June 390 were held (M4 and D5), so M3 and D4
must have taken place in 391.
It is strange that Xenophon does not mention the Athenian armys resistance to
Agesilaus attack, when the Athenians came en masse to Corinth to fortify the Long
Walls, fearful of the Lacedaemonians reaching Attica. The most convincing explanation is that the Athenian army had withdrawn between the reconstruction of the
Long Walls and Agesilaus expedition, after the Corinthian Walls had been rebuilt.
This withdrawal must relate to the demobilization at the end of the campaigning
season, but Xenophon does not mention it for the simple reason that it affected only
the Athenians and not the Lacedaemonian army, which had not been mobilized. This
28
We do not need to assume that Praxitas failed to garrison the Lechaeum in 393. It would be
strange for Praxitas to station contingents in Sidus and Crommyon and not at Lechaeum,
however exposed the position was. This suggests that Hell. 4.4.17 is out of chronological context
and refers to 4.4.713. According to Xenophon the attack on the Lechaeum garrison took place
between Iphicrates campaigns and the reconstruction of the walls by the Athenians, which
makes sense. Diodorus account (14.91.2), which mentions an attack by the Corinthian exiles
from the Lechaeum before Iphicrates actions in Phleious and Sicyon, can be placed in the same
context. As a result of these attacks the Lacedaemonians and their allies had to evacuate
Lechaeum in 392 after the fighting for the Athenians to take advantage of the situation and
rebuild the Long Walls.

82

JOS PASCUAL

leads us to various conclusions: during 393 and 392 the Lacedaemonian army was not
mobilized to invade Corinth and the fighting was done by a Lacedaemonian mora
and allied contingents or mercenaries; and in two cases, Hell. 4.4.1 and 4.4.14, the
implies the beginning of a year.
expression
Thus Praxitas campaign took place in 393, Iphicrates incursions, the capture of
Lechaeum by the allies and the reconstruction of the Corinthian Long Walls in 392
and Agesilaus expedition against Argolid and Corinth in 391; finally the mobilization
of Hell .4.5.1, M4, Agesilaus second expedition into Corinthian territory, during the
course of which he took the Corinthian Peiraion, the Isthmian Games and the
disaster of the Lacedaemonian mora all took place in 390. Xenophon refers to the
demobilization of the Lacedaemonian army in that year, around July/August, in Hell.
4.5.18 (D5).
In the context of early 393, Hell. 4.4.1 can easily be interpreted as a reference to the
various garrisons established by the respective contenders in Corinth and Sicyon.
Thus the Lacedaemonian mora stationed in Sicyon is mentioned on a number of
occasions (Hell. 4.4.7, 8, 17, 4.5.11), and there was also a Boeotian garrison in
Lechaeum (4.4.9, 12); the Athenian mercenary garrison is well attested,29 and the
Argive presence (4.4.9, 10, 11, 13), despite Andocides rhetorical exaggeration (3.18),
must relate to the garrison in Corinth, which was forced to leave the city after the
Kings Peace was signed (5.1.34, 36). Xenophon consistently fails to mention
Phleious, which only accepted a garrison at the end of the year, after the invasion by
Iphicrates and his peltasts (4.4.15).
In Hell. 4.4.14 Xenophon asserts that from the beginning of 392, the contenders no
). Xenophon may be
longer deployed great armies (
linking and relating this development with his immediately preceding reference to the
demobilization of the Lacedaemonians allies (Sicyon, Phleious, etc.) (4.4.13), whose
armies were mobilized en masse in 393 after the victory of the Long Walls (the
of that year). Before the end of 392, the Athenians also came en
masse to rebuild the Long Walls.
We do not need a whole year, 393, to elapse for Corinthian discontent at the
plundering of its territory (cf. Hell. 4.4.2) to erupt violently in the Eucleiae of 392.
Demosthenes testimony suggests that stasis in Corinth had already begun at the time of
the battle of Nemea (July 394). Thus, immediately after the battle of Nemea, Demosthenes asserts (20.523) that some of the Corinthians tried to prevent the allied army,
which was returning defeated from Nemea, from taking refuge behind the city walls.
Dating the battle of the Corinthian Long Walls to 392 leaves little time for
Andocides IIIs speech, which refers to that battle (3.18), between the Lacedaemonian
occupation of Lechaeum and the reconquest by the allies in the same Julian year.
The speech may have been given in the archonship of Philocles, Attic year 392/1.
Andocides (3.20) mentions the four years of war that would correspond with the
campaigning season of 392 and Attic year 392/1. These dates are supported by the
fragment of Philochorus (FGrH 328 F149 a), recorded by Didymus, in which
the meeting in Sparta to consider peace falls within the archonship of Philocles and
refers to the beginning of 392/1.30 This suggests that the peace
the entry

29

Xen. Hell. 4.4.9, 1517. For the other references, see Harding (n. 12), no. 22, 3538.
See Didymus in Dem. 10.34, col. 7.1114. Xenophon does not mention the Congress of
Sparta in 392/1 and Plutarch (Ages. 23.14) confuses successive missions by Antalcidas (see E.
Aucello, La genesi della Pace di Antalcida, Helikon 5 [1965], 364).
30

XENOPHON AND THE CHRONOLOGY OF THE WAR

83

negotiations began in Sardis in the spring of 39231 and broke down in Sparta before
Andocides gave his speech in Athens, at the beginning of a new Attic year, and before
the reconstruction of the Corinthian Long Walls in August/September of 392. So, in
the context of the peace negotiations of 392, it is better dating the battle of the
Corinthian Long Walls to 393 (July/August) and not to 392 (July/August).
It is worth asking why the Lacedaemonian army did not come to the Isthmus in the
years 393 and 392. In both cases we can make a connection between the war on land
and the events that took place at sea in 393 and 392. After the naval battle of Cnidus
in the autumn of 394, Conon and Pharnabazus sailed north with the objective of
expelling the Lacedaemonian garrisons from the Greek cities of Asia Minor and the
islands near the coast (Xen. Hell. 4.8.1).32 Diodorus says that first they persuaded Cos
to rebel, then Nisyros and finally Teos.33 Afterwards, Conon and Pharnabazus set sail
for Chios where the Chiotes expelled the Lacedaemonian garrison and joined Conon.
After Chios, Erythrae, Mitilene, Ephesus34 and perhaps Samos35 also went over to
Conon.
Once the fleet reached Ephesus, Pharnabazus disembarked and went overland to
his satrapy.36 Conon, with forty ships, continued up the coast to Sestus, where he once
again joined Pharnabazus. Along the route Conon must have won over practically all
the cities that had been allies of the Lacedaemonians with the exception of at least
Sestus and Abydos, which remained under the control of the Spartan Dercylidas.37
After their rendezvous on the coast close to Sestus, Pharnabazus returned to his
residence and Conon wintered in the Hellespont, with the intention of winning over
the Greek cities in the area and preparing a great naval expedition in the Aegean with
Pharnabazus the following spring.38
Although Diodorus (14.84.35) includes Conon and Pharnabazus campaign on
the Asian coast and the subsequent expedition across the Aegean to Laconia and
Corinth in a single year, we can be fairly safe in assuming that the Aegean expedition
was mounted in 393.39 In fact on the only occasion when Xenophon refers to
campaigning seasons in the naval war (Hell. 4.8.7), he expressly says that Conon and
Pharnabazus spent the winter in the Hellespont and, when spring arrived, after
manning many ships and hiring a force of mercenaries, Pharnabazus sailed through
the islands to Melos, which he and Conon used as a base to attack Lacedaemonia (
31 J.G. DeVoto, Agesilaus, Antalacidas, and the Failed Peace of 392/1 B.C., CPh 81 (1986),
191202.
32 Funke (n. 5), 82; Seager (n. 5), 101; R.J. Buck, Thrasybulus and the Athenian Democracy
(Stuttgart, 1998), 105; Buckler (n. 5), 130.
33 Only in the case of the island of Teos, situated on the northern coast of Ionia, between
Ephesus and Erythrae, is the geographical order of Diodorus description apparently broken.
However we can assume that, instead of
or
, this should read
, Telos, the
small island situated opposite the Cnidian Peninsula. Nisyros is the closest island to the south,
and Telos is halfway between Cos and Rhodes (L. Robert, Diodore, XIV, 84, 3, Revue de
Philologie 8 [1934], 438).
34 Diod. 14.84.3. Statues were erected in honour of Conon in Erythrae, Ephesus and Samos (cf.
Paus. 6.3.16; Seager [n. 5], 102; Tod [n. 11], no. 106; Harding [n. 12], no. 12 D; Rhodes, Osborne
[n. 11], no. 8, 447).
35 G. Shipley, A History of Samos 800188 B.C. (Oxford, 1987), 134.
36 Xen. Hell .4.8.3.
37 Xen. Hell. 4.8.35.
38 Xen. Hell. 4.8.67; Funke (n. 5), 82; Seager (n. 5), 103; Buck (n. 33), 105, 107; Buckler (n. 5),
1312.
39 Beloch (n. 3), 778; Judeich (n. 4), 143; Cartledge (n. 5), 362; Seager (n. 5), 104; Buck (n. 32),
103; Buckler (n. 5), 1345.

84

JOS PASCUAL

).
Thus, in the spring of 393, probably around March/April, Conon and Pharnabazus
set sail for the Cyclades. When they arrived, Melos, Delos, Paros and probably the rest
of the Cyclades defected from their alliance with the Lacedaemonians. The crowns
offered by the Athenians at the Great Delia, held every four years, probably indicate
the time when Spartan influence in Delos ended as a result of Conon and Pharnabazus expedition. These Athenian crowns were dedicated in the late-fifth-century
temple at Delos known as the temple of the Athenians or the temple of the Seven
Statues.40 There is a record of twenty-one crowns offered between 417 and 329. Of
these, the fourth belongs to 405 and the fifth to 393. The crown for 401 is missing,
evidently because the island was at that time under Spartan control.41 Similarly, the
crown offered by the Athenians at the Delia of 393 indicates the terminus ante quem
for Delos abandoning the Spartan alliance. The Great Delia were held in spring,
during the month of Thargelion (May/June) in the third year of each Olympiad, and
the fifth crown proves that Conon had, around the end of May or more probably the
beginning of June 393, not only wrested Delos from Spartan control, but had by then
reached Athens. Significantly the crown of 393 was valued at 122 drachmas,42 and was
the richest ever offered by Athens at Delos, which could be connected with the Persian
money given by Conon to the Athenians.
From Melos, Pharnabazus and Conon sailed towards the southern coast of the
Peloponnese. They occupied Pherae or Pharae, modern Kalamata, on the mouth of
the river Nedos43 and laid waste to its territory and other places along the coast. Their
plan to occupy Pharae, on the Messenian coast, was probably aimed at stirring up a
revolt of the helots, just as Pylos had done during the Peloponnesian War, and also to
establish a base from which they could attack Spartan territory directly.44 However,
according to Xenophon (Hell. 4.8.7), the lack of good harbours in the area, shortage
of food supplies and the proximity of relief forces forced Conon and Pharnabazus to
withdraw (
). The shortage of food in fertile Messenia may indicate that the wheat
was still green,45 that is, Conon must have reached Messenia before the middle of
May.
40 F. Courby, Notes topographiques et chronologiques sur le sanctuaire dApollon Dlien,
BCH 45 (1921), 179. Seager (n. 5), 105 says that Athens controlled the amphictiony of Delos in
393/2, but does not connect this with Conon.
41 An inscription recalls this Spartan control of Delos, which Tod (n. 12), no. 99, 67 dates to
around 403. The inscription cannot be before Aegospotami (405) and refers to kings Agis and
Pausanias and the Council of the ephors. Agis died around the summer of 400 and one of the ephors
of 403 that we know of, Naucleidas (Xen. Hell. 2.4.36), is not mentioned in this inscription, and
neither is Sciraphidas or Phlogidas (Plut. Lys. 17), who was also probably an ephor in 404. The most
plausible date is thus 402 or 401, perhaps during the Delia of 401.
42 Courby (n. 40), 1804.
43 Strab. 8.4.5; Paus. 4.1.4, 4.2.10, 4.16.8, 4.30.26, 4.3.31 (Pharae is six stadia from the sea, about
one kilometre). Cf. R. Baladi, Le Ploponnse de Strabon. tude de gographie historique (Paris,
1980), 69, 23940.
44 Thuc. 4.34, 5.56.3, 5.115.2, 6.105.2, 7.18.3, 26.
45 For a similar situation see Thuc. 4.6.1: in 425 the Peloponnesians who invaded Attica early in
the season when the wheat was still green were short of food.

XENOPHON AND THE CHRONOLOGY OF THE WAR

85

From Pharae, without giving up the idea of creating an anti-Spartan base, the
Persian fleet turned back and went to Phoenicus on Cythera, which was well situated
for attacking Laconia and interrupting Laconian trade,46 and whose walls were in a
bad state of repair. When they saw the Persian fleet coming, the Cythereans surrendered without a fight.47 Conon and Pharnabazus repaired the walls of Phoenicus
and left a garrison there, under the command of Nicophemus, and then sailed along
the Peloponnesian coast to Corinth,48 where the Persian fleet probably anchored at
the end of May.
Once in Corinth, Conon and Pharnabazus met the Council of the allies, not before
May/June 393.49 Pharnabazus gave them money and lifted their spirits, which had
been severely undermined by the defeats of 394 in Nemea and Coronea and the
pillage of Corinthian territory. He then returned to Asia, but before he left, Conon
persuaded him to help reconstruct the Athenian walls. Thus, around the beginning of
June, Conon entered Piraeus with eighty of the Persian fleets ships and provided the
necessary funds for continuing work on the Athenian walls,50 which had begun some
considerable time before, and for which the Athenians offered a rich crown in the
Delia.
In Funkes opinion no major campaign took place on land in 393 and the war was
probably restricted to a war of manoeuvre in which Corinth was particularly hard
hit.51 He attributes this to the loss of Spartas naval supremacy as a result of the battle
of Cnidus, which meant it did not want to get involved in major military operations
that year, and Cartledge assumes that Pharnabazus and Conons naval campaign must
have caused some kind of panic in Sparta, although Xenophon does not mention it.52
Here we can perhaps identify the forces mentioned by Xenophon which came to
Messenias defence in 393. The Lacedaemonians had one of their six morai stationed
in Orchomenus and another in Sicyon, but a large contingent of troops must have
been approaching, because the powerful Persian fleet had to set sail with a large force
of mercenaries. Thus we can assume that the relief forces actually consisted of the
Lacedaemonian army, which had been mobilized to face the invasion.53
The scenario we have described can be compared with similar situations that are
better known. In 425, when the Lacedaemonians discovered that the Athenians had
taken Pylos, they ordered the army that had invaded Attica to return (Thuc. 4.6.12).

46 Xen. Hell. 4.8.8. The occupation of Cythera formed part of a long-standing strategy of harrying Sparta. On this and the islands importance both in the Persian Wars and in the Peloponnesian
War, see Hdt. 7.235.14; Thuc. 4.535.
47 Cf. P. Karavites, Capitulations and Greek Interstate Relations. The Reflection of Humanistic
Ideals in Political Events (Gottingen, 1982), 77.
48 Xen. Hell. 4.8.78.
49 Funke (n. 5), 83.
50 IG 22 1656 and 1657; Tod (n. 11), nos 107 A and B; Harding (n. 12), no. 17, 37; Rhodes, Osborne
(n. 11), no. 9, 469; Xen. Hell. 4.8.911; Lysias 2.63; Diod. 14.85.24; Nepos, Conon. 4.5; H.
Swoboda (1922), Konon, RE 11, 1331; G. Barbieri, Conone (Rome, 1955), 1678; Underhill (n. 8),
xlvi. The reconstruction of the walls probably began immediately after the battle of Haliartus
(S. Perlman, Athenian democracy and the revival of imperialistic expansion at the beginning of
the fourth century, CPh 63 [1968], 261; Seager [n. 5], 101; Buck [n. 32], 99). In any case, thanks to
the inscriptions we know that the work began during the archonship of Diophantus (395/4), before
the arrival of Conon, but Persian money greatly speeded it up.
51 Funke (n. 5), 81.
52 Cartledge (n. 5), 362.
53 Isocrates (4.119) also refers to the havoc wreaked by the Persian fleet on the Peloponnesian
coast.

86

JOS PASCUAL

In the spring of 375, when the Lacedaemonians were preparing to invade Boeotia, the
Boeotians asked the Athenians to sail around the Peloponnesian coast, because they
reasoned that the Lacedaemonians would be unable to guard their own territory, the
allied cities and send a sufficiently large army against them all at the same time (Xen.
Hell. 5.4.62). The Athenians sent Timoteus and, sure enough, the Lacedaemonians
did not invade Thebes that year because the Athenian fleet was close to their coasts
(Hell. 5.4.63).
A similar situation probably developed in 393. It seems unlikely that the Lacedaemonians, with two morai away and at least part of the army committed to defending
the Spartan coasts, could have mobilized the army to march against Corinth, all in the
same year. The Lacedaemonians were obviously aware that the Persian fleet was in the
Aegean before the beginning of the campaigning season on land in April/May, and it
was very dangerous to march to the Isthmus leaving Laconia and Messenia ungarrisoned. Thus the threat from Conon and Pharnabazus prevented them mobilizing
the Lacedaemonian army at the beginning of 393 and Conons subsequent attack
once again prevented them sending the army to Corinth after the battle of the Long
Walls, after which only the allies mobilized.
Consequently, Praxitas demobilization (Hell. 4.4.13) must have been in 393. At the
beginning of that year, therefore, with no mobilization of the Lacedaemonian army,
the Corinthian oligarchs were presumably massacred around March 393, to judge
from Diodorus account (14.86.56);54 Conon arrived around May/June and stationed
a mercenary force in Corinth, under the command of Iphicrates, and once the
situation had become calmer, the exiles returned. Since Iphicrates mercenaries took
part in the battle of the Long Walls (Hell. 4.4.9), Praxitas must have achieved his
victory after Conon reached Corinth. Argos and Corinth united after Conons march,
probably as a sympoliteia,55 and the battle of the long Walls was fought and
Lechaeum taken in the second half of the year, around August/September. The allies,
Sicyonians, Phliasians, Arcadians, etc., joined Praxitas en masse after the battle. At
the end of the (Lacedaemonian) year Praxitas took Sidus and Crommyon and
disbanded the army in September 393.
After Conon withdrew, a threat remained from the enemy presence in Cythera,
from the Persian fleet anchored in the Piraeus and from the revitalization of the
Corinthian fleet. Just as the Lacedaemonian army did not invade Attica the following
year, 424, because of the occupation of Pylos, it would not have been mobilized to
fight in Corinth in 392. That year Sparta was busy defending its coasts and gaining
naval supremacy, as we shall see, in the Gulf of Corinth.

54 Cf. Judeich (n. 4), 143 who also dates the battle of the Corinthian Long Walls to 393.
Aucello (n. 8), 378 believes that the massacre of Corinth took place in March 393 and the Isthmian
Games and the destruction of the mora in the archonship of Demostratus in 392 (Attic year 393/2).
Gioia (n. 25), 38 considers 393 more likely for the massacre of the oligarchs, apart from which
Diodorus includes it in the Attic year 394/3. For a date of the massacre of the Corinthian oligarchs in
March 392 and the taking of Lechaeum in the same year, see Grote (n. 3), 152; Underhill (n. 8),
xlviii and n. 1 (although he dates it in February); Griffith (n. 7), 2412, 250; D. Kagan, Corinthian politics and the revolution of 392 B.C., Historia 11 (1962), 44757; C.D. Hamilton, Spartas
Bitter Victories. Politics and Diplomacy in the Corinthian War (Ithaca and London, 1979), 249,
267; Funke (n. 5), 82, 85, 87; M. Whitby, The union of Corinth and Argos, Historia 33 (1984),
2957; Cartledge (n. 5), 363; Seager (n. 5), 106.
55 Hamilton (n. 5), 113.

XENOPHON AND THE CHRONOLOGY OF THE WAR

87

The fighting between Corinthians and Lacedaemonians in the Gulf of Corinth

provides us with further chronological information (Hell. 4.8.1011).56 As we saw,
around May/June 393 Pharnabazus and Conon reached Corinth with a Persian fleet.
The Council of the allies then met and Pharnabazus provided additional funds to
wage the war. The Corinthians used Pharnabazus money to build a squadron and
gain hegemony over the Gulf of Corinth. The Spartans also armed a fleet. The
Corinthians won a naval battle in which Podanemus, probably the Spartan nauarchos
of 394/3, was killed. The naval battle could have taken place in the middle of summer
393, before August/September, and presupposes that the Corinthians still controlled
Lechaeum. The Corinthian nauarchos Agathinus, in reality one of the strategoi,57 was
subsequently relieved of command by Proaenus, at the time of the Corinthian new
year. The new Corinthian commander abandoned Rhium, which was taken by the
Lacedaemonians. We can assume that the Corinthian fleet was probably at Cape
Rhium because Lechaeum was occupied by the Lacedaemonians (c. August/
September 393) and may have withdrawn from Rhium because the allies had retaken
Lechaeum (c. August/September 392). Then Teleutias arrived, perhaps in October
392, and gained naval control of the Gulf in the spring of 391; this could already have
happened before Lechaeum was occupied by the Lacedaemonians for a second time in
391.
Furthermore, the peace negotiations in Sardis no doubt took place the same spring
of 392 and opened up reasonable possibilities of agreeing on a peace treaty to settle
the conflict.58 For all these reasons the Lacedaemonians did not mobilize the army in
392 and so there were only skirmishes between the garrisons of Sicyon and Corinth in
the first half of 392 (Hell. 4.4.14); Iphicrates incursions into Phleious and Arcadia
would have occurred in May and June when the wheat was ripe, and we can calculate
that Lechaeum was retaken and the Corinthian Long Walls rebuilt around August/
September (Hell. 4.4.1518). Once this had been done the Athenian army returned
from Corinth, considering the year finished.
The Spartan occupation of Rhium (392), her supremacy in the Gulf (spring of
391), the breakdown of peace negotiations and Conons arrest by Tiribazus (Hell.
4.8.16) must have finally lifted the threat to the Lacedaemonian coasts and made it
possible to mobilize the Lacedaemonian army in 391. Agesilaus campaign in the
Argolid would have taken place at the beginning of the campaigning season of 391
(M3), in the archonship of Nicoteles, between April and May,59 and Lechaeum was
probably taken for the second time in a coordinated amphibian attack led by Teleutias
at sea and Agesilaus himself on land, in June or July. The occupation of Lechaeum
consolidated Spartan hegemony in the Gulf and Teleutias departed for Cnidus and
Rhodes in the second half of the year.60 After the conquest of Lechaeum, Agesilaus
disbanded the army, thus ending the 391 campaigning season (Hell. 4.4.19; D4).
Despite the Spartan expedition of 391, the Corinthians succeeded in obtaining
supplies through the Boeotian port of Creusis and cultivated Peiraion, an area to the
56 Funke (n. 5), 835; Underhill (n. 8), xlvi; Seager (n. 5), 105. Hamilton (n. 55), 231 dates the
naval operations in the Gulf to the latter part of 393.
57 Xenophon (Hell. 4.8.10) applies Spartan terminology to Corinthian magistrates, who did
not have a nauarcheia.
58 Beloch (n. 3), 812 and Funke (n. 5), 856 for whom Argos refusal presupposes control of
Corinth, since its possession was of strategic importance for prosecuting the war. Hamilton
(n. 5), 111 dates them in spring.
59 For a date in 392 for Agesilaus campaign, cf. Judeich (n. 4), 1456.
60 Funke (n. 5), 84; Buck (n. 32), 112.

88

JOS PASCUAL

north of the Isthmus far from the Spartan garrisons. In 390 the Lacedaemonians
undertook another expedition against Corinth under the command of Agesilaus. The
Spartan king probably left with the army around April or May, celebrated the
Isthmian games of May/June 390 and took Corinthian Peiraion, probably in June
(Hell. 4.5.16). In the same period, around June, just as the Hyacinthia were about to
be celebrated in Sparta, the disaster of the Lacedaemonian mora occurred (Hell.
4.5.718),61 after which Agesilaus suddenly returned to Lacedaemonia (probably in
July; Hell. 4.5.1819).62 After Agesilaus withdrawal in the second half of the year,
Iphicrates took advantage of the Lacedaemonian defeat to occupy Sidus, Crommyon
and Oenoe and imprison the Lacedaemonians in Lechaeum (probably August/
September).63 In the winter of 390/89 friction developed between Iphicrates and the
Argives in Corinth, as a result of which Iphicrates was replaced, in the spring or
summer of 389, by Chabrias and sent to Thrace the following year, in the spring of
388.64
We know of two demobilizations in 389, 388 or 387 (D5 and D6), the first of which
relates to an expedition by Agesilaus in Acarnania (M5) and the second to an
expedition by Agesipolis against Argos (M7). The Achaean envoys very possibly
arrived in Sparta at the beginning of spring 389 to ask for its help against the
Acarnanians. The Lacedaemonians ordered the expedition (M5, Hell. 4.6.1), and it
was undertaken in the summer of 389 by Agesilaus, who returned in the autumn
(Hell. 4.6.12), possibly in September 389 (D6).65
The Acarnanians surrendered at the beginning of spring 388, under the threat of
another invasion (M6, Hell. 4.7.1); this left the Spartan army free to undertake an
expedition against Argos under the command of Agesipolis in the summer of 388
(M7 and D7, Hell. 4.7.27). However, there is another possibility. When the
Acarnanians arrived to negotiate the peace, the ephors had already decreed the
mobilization for the following year (Hell. 4.7.1) and they very possibly entrusted
Agesilaus with commanding the expedition. Agesilaus had in fact promised the
Achaeans that he would return the next year to Acarnania (Hell. 4.6.13), but the
expedition against Argos was commanded by Agesipolis. While the army was massing
in Phleious in order to attack through Nemea, Agesipolis consulted first Zeus and
then Delphi on whether the sacred truce sought by the Argives was a just one, which
took some time. Xenophon (Hell. 4.7.2) introduces Agesipolis expedition with the
which he had already used on two other occasions, as we have
formula
seen, to indicate that a new year had begun. So this determines when the activities of
Antalcidas, the fighting against the Athenian fleet and the beginning of the negotiations for signing the Kings Peace took place.
Antalcidas was nauarchos in 388/766 and departed for the Kings court in Susa in
the autumn/winter of 388. In the summer of 387 Antalcidas and Tiribazus were once
61 Beloch (n. 3), 21920; Griffith (n. 7), 246; P. Cloch, Thbes de Botie. Des origines la
conqute romaine (Namur, 1952), 111; Hamilton (n. 54), 281; P. Cartledge, Sparta and Lakonia. A
Regional History 1300362 B.C. (London, Boston and Henley, 1979), 286; Funke (n. 5), 91; Whitby
(n. 55), 297.
62 The campaign is usually dated to the spring, cf. Beloch (n. 3), 86; Underhill (n. 8), xlix;
Funke (n. 5), 91; Hamilton (n. 5), 114; Seager (n. 5), 110; Buckler (n. 5), 11718.
63 Funke (n. 5), 92; Buckler (n. 5), 121.
64 Schol. Aristeides Pan. 274 f D; Polyaen. 3.11.6, 15; Griffith (n. 7), 243; Seager (n. 5), 111.
65 Beloch (n. 3), 86; Funke (n. 5), 92; Hamilton (n. 5), 116; Seager (n. 5), 111; Buckler (n. 5),
1257.
66 Xen. Hell .5.1.6; Underhill (n. 24), 136.

XENOPHON AND THE CHRONOLOGY OF THE WAR

89

again in Asia Minor and for the rest of the summer, before the end of Antalcidas
nauarcheia in September, were no doubt occupied in gaining naval hegemony. In the
autumn in Sardis Tiribazus learnt of the Kings conditions for a peace treaty.
Xenophon (Hell. 5.1.29) says that the Athenians were in favour of peace because of
their naval defeat at the Hellespont, and he also links the Argive desire to sign the
peace to the invasion of their land, and, we can assume, Agesipolis campaign in the
Argolid. In short, Agesipolis expedition does not appear to be related to the mobilization ordered by the ephors in the spring of 388 and it took place not long before the
peace negotiations, so it could set out in 387.
Peace negotiations continued throughout 387 and the peace treaty was signed in
Sparta in the spring of 386.67 During these peace negotiations, Agesilaus persuaded
the ephors to mobilize the army because the Thebans had refused to dissolve the
Boeotian Confederacy (Hell. 5.1.33). He crossed the frontier of the Lacedaemonian
state and began to muster the allies in Tegea. The Thebans arrived while he was there
and acceded to the Spartan demands, so there was another mobilization in the
spring/summer of 386 (M8).
Once the Thebans had accepted the terms of the Peace, Agesilaus ordered the
Corinthians and Argives to force the Argive garrison to leave Corinth (Hell. 5.1.34)
and stopped the Argives taking over Corinth by threatening to make war on them if
they did not leave the city (Hell. 5.1.35). Xenophon may be referring to another
mobilization ordered by the ephors against Argos that same year. However, it was not
necessary to demobilize the army that was already stationed in Tegea first. Once the
Thebans agreed to peace, the ephors, who were empowered not only to order mobilization but also to decide where military operations were to take place, only had
publicly to announce the change of objective. The final demobilization in 386, in
practice only that of the Lacedaemonian army and its allies, is also recorded by
Xenophon (Hell. 5.1.35; D8).
Our conclusions can be summarized as follows (see Table 2).
Universidad Autnoma de Madrid

67

Polyb. 1.6; Arist. 2.370; Underhill (n. 8), l.

JOS PASCUAL
jose.pascual@uam.es

90

JOS PASCUAL

Table 2
Mobilizations
Year 395
M1. Xen. Hell. 3.5.6. The ephors order
mobilization.

Year 394
M2. Xen. Hell. 4.2.9. The ephors order
mobilization.
Year 393
(1), Xen. Hell. 4.4.1.
[Beginning of a year]

Demobilizations

D1. Xen. Hell. 3.5.225. Withdrawal of the

Lacedaemonian army after the battle of
Haliartus.

D2. Xen. Hell. 4.4.1. Agesilaus disbands the

army.
D3. Xen. Hell. 4.4.13. Praxitas,
Lacedaemonian harmost in Sicyon, disbands
the army.

Year 392
(2), Xen. Hell. 4.4.14.
[Beginning of a year]
Xen. Hell. 4.4.18. [End of a year]
Year 391
M3. Xen. Hell. 4.4.19. The Lacedaemonians
undertake an expedition against Argos (and
Corinth).
Year 390
M4. Xen. Hell. 4.5.1. The Lacedaemonians
undertake another expedition against
Corinthian territory.

D4. Xen. Hell. 4.4.19. Agesilaus disbands the

army.

D5. Xen. Hell. 4.5.18. Agesilaus returns to

Lacedaemonia with the army.

Year 389
M5. Xen. Hell. 4.6.13. Mobilization against
the Acarnanians.
D6. Xen. Hell. 4.6.12, 14. When autumn
arrives, Agesilaus withdraws from Acarnania.
Year 388
M6. Xen. Hell. 4.7.1. Mobilization against the
Acarnanians, who capitulate before the
expedition.
Year 387
M7. Xen. Hell. 4.7.2. Mobilization against the
Argives.
D7. Xen. Hell. 4.7.7. Agesipolis, the Spartan
king, disbands the army after a campaign in
the Argolid.
Year 386
M8. Xen. Hell. 5.1.33. Agesilaus persuaded
the ephors to order the mobilization (against
Thebans, after against Argives and
Corinthians).

D8. Xen. Hell. 5.1.35. General demobilization

of the army and navy

Classical Quarterly 59.1 91111 (2009) Printed in Great Britain

doi:10.1017/S0009838809000007X

91
AN UNIDENTIFIED FRAGMENT PHILIP
OF DIODORUS
RANCE

HANNIBAL, ELEPHANTS AND TURRETS IN

SUDA
438 [POLYBIUS FR. 162 B ] AN
UNIDENTIFIED FRAGMENT OF DIODORUS*
Suda

438:

.1
Thrakion: Hannibal, the Carthaginian general, by carrying the turrets of the elephants and
making use of the beasts litters to cut away the branches to the greatest height, made the route
secure and easy.

This entry in the Suda has had a chequered history as a potential fragment of
Polybius Historiae. It was first identified as Polybian by Casaubon and included in
the editio princeps of 1609, and was retained in subsequent editions, though one
suspects owing more to scholarly inertia than conviction. It was excised by
Schweighuser (178995) and restored by Bttner-Wobst (1904), but marked in
square parenthesis as dubious. It thus resides in the current Teubner edition among
the fragmenta ex incertis libris [162b].2 It continues to be cited in both general and
specialist lexica as a doubtful or spurious fragment of Polybius.3 In consequence its
authenticity has been universally distrusted in modern scholarship from various
perspectives lexical, textual and historical: in her edition of the Suda Adler marks
438 as a dubious citation of Polybius; in his classic study of the elephant in
* I am grateful to Nicholas Sekunda (University of Gdansk) and Michael Charles (Southern
Cross University NSW) for their comments, and to the latter for kindly allowing me to read his
paper (2008) in advance of publication. The following abbreviations have been used: Bigwood
(1980) = J.M. Bigwood, Diodorus and Ctesias, Phoenix 34.3 (1980), 195207; Chamoux and
Bertrac (1993) = F. Chamoux and P. Bertrac, Diodore de Sicile: Bibiothque Historique. Introduction gnrale. Livre I ([Bud] Paris, 1993); Charles (2008) = M.B. Charles, African forest
elephants and turrets in the Ancient World, Phoenix, forthcoming; Charles and Rhodan (2007)
= M.B. Charles and P. Rhodan, Magister Elephantorum: a reappraisal of Hannibals use of
elephants, CW 100.4 (2007), 36389; Goukowsky (1972) = P. Goukowsky, Le roi Prus, son
lphant et quelques autres, BCH 96 (1972), 473502; Goukowsky, Diodore XVII = P.
Goukowsky (ed. and French trans.), Diodore de Sicile, Bibliothque Historique Livre XVII ([Bud]
Paris, 1976); Goukowsky, Fragments II = P. Goukowsky (ed., French trans. and comm.), Diodore
de Sicile, Bibliothque Historique. Fragments II. Livres XXIXXVI ([Bud] Paris 2006); Gowers
and Scullard (1950) = W. Gowers and H.H. Scullard, Hannibals elephants again, NC (ser. 6) 10
(1950), 27183; Scullard (1948) = H.H. Scullard, Hannibals elephants, NC (ser. 6) 8 (1948),
15868; Scullard (1974) = H.H. Scullard, The Elephant in the Greek and Roman World (London,
1974); Walbank, Commentary = F.W. Walbank, A Historical Commentary on Polybius (Oxford,
195767).
1 A. Adler (ed.), Suidae Lexicon (Leipzig, 192838; repr. Stuttgart, 196771), 2.724.
2 I. Casaubon (ed.), Polybii Historiarum libri qui supersunt (Paris, 1609), 1027; J. Schweighuser (ed.), Polybii Megalopolitani Historiarum quidquid superest (Leipzig, 178995), 5.601, fr.
22; T. Bttner-Wobst (ed.), Polybius, Historiae ([Bibliotheca Teubneriana] Leipzig, 1904; repr.
Stuttgart, 1967), 4.535, fr. 162b.
3 E.g. H. Stephanus et al. (edd.), Thesaurus Graecae Linguae (Paris, 183165), 4.474, s.v.
, 5.1781, s.v.
; LSJ9, 813, s.v.
, 1203, s.v.
[Plb.] fr. 162b;
A. Mauersberger (ed.), PolybiosLexicon (Berlin, 195668; repr. 2006), Bd 1.3, 1164, s.v.
, Bd. 1.4, 1722, s.v.
: [fr. 162b].

92

PHILIP RANCE

Graeco-Roman antiquity Scullard dismisses the entry on the grounds that the source
is not known and only very doubtfully could it be attributed to Polybius; while
Walbank excludes the fragment from his commentary.4 To the present authors
knowledge, however, Polybius authorship has never been defended or refuted with
argumentation. If it is true that the ascription to Polybius was never more than
Casaubons overconfident surmise, can the fragment be assigned with confidence to
another author? There is more at stake than a purely parochial issue of Polybian
studies or locating a new abode for a homeless fragment. The wider historical significance of Suda 438 for the campaigns of Hannibal is that, excepting some
demonstrably fictive allusions in later Latin poetry, this fragment contains the only
explicit and unequivocal statement that Hannibals elephants were furnished with
turrets. Furthermore, all previous studies of this armament have overlooked the
important information concerning its terminology and construction that this fragment uniquely preserves. Identification of the author and ultimate source will
therefore contribute to the protracted debate over whether African forest elephants,
distinct from the Indian species, were so equipped in Carthaginian armies or more
broadly in the Hellenistic period.
The first task in the process of identification is to narrow the field of candidates. It
has long been recognized, if not widely appreciated, that the compiler of the Suda
(c. 1000) did not derive its numerous quotations of historical authors from first-hand
acquaintance with the original works but rather though the convenient medium of a
recently published encyclopaedia. De Boor demonstrated that all the historical entries
in the Suda, with the exception of those that originate from earlier lexica or scholia,
were drawn from the Excerpta Constantiniana, the thematic volumes of historical
extracts compiled at the direction of Constantine VII Porphyrogenitus (91359), with
the overwhelming majority of citations being, as here, anonymous.5 There is no
reason to believe that Suda 438 uniquely derives from a different source or methodology.6 The compiler of the Suda used only a small number of volumes of Excerpta.
Of those volumes fully or partly preserved today, he made use of De virtutibus et vitiis
and De legationibus, but not De insidiis or De sententiis. In addition, the Suda drew
material from volumes that have not survived, among which De Boor identified
the second half of De virtutibus, as well as the
4 A. Adler (ed.), Suidae Lexicon, loc. cit. (app. crit.); Scullard (1974), 242 (reiterating id. [1948],
162, n. 9); Walbank, Commentary, III 753.
5 The dependence of the Suda on the Excerpta Constantiniana was first observed by H.
Valesius (ed.), Polybii, Diodori Siculi Excerpta ex Collecteneis Constantini Augusti
Porphyrogentae (Paris, 1634), unpag. pref., but the nature and extent of the relationship was
demonstrated by C. de Boor, Suidas und die Konstantinische Exzerptsammlung, BZ [pt. I] 21
(1912), 381424; [pt. II] 23 (191419), 1127; and exemplified in previous case studies: Zu
Iohannes Antiochenus, Hermes 20 (1885), 32130, at 32730; Die Chronik des Georgius
Monachus als Quelle des Suidas, Hermes 21 (1886), 126. See also J. Becker, De Suidae Excerptis
Historicis (Bonn, 1915); A. Adler (ed.), Suidae Lexicon, I xixxi; reiterated in id., Suidas
(Lexikograph), RE 4A.1 (1932), 675717, at 679, 7006; P. Lemerle, Le Premier Humanisme
Byzantin. Notes et remarques sur enseignement et culture Byzance des origines au Xe sicle (Paris,
1971), 2857.
6 There is no question that Suda
438 might derive from an earlier lexicon and scholion; De
Boor (n. 5 [191419]), 2237 shows that in historical entries originating from such sources the
author is always cited, mostly using
or
. In contrast, the majority of entries deriving
from the Excerpta Constantiniana are anonymous; in the minority of instances where the author
is specified he is almost always named in the nominitive preceding the citation, occasionally with
or
.

AN UNIDENTIFIED FRAGMENT OF DIODORUS

93

, and perhaps one other volume with

military content. The Suda therefore becomes an indirect source for the contents of
these lost volumes.7 Since Suda 438 is not found in the extant volumes of Excerpta,
it must come from a historical work that was excerpted in one of the lost volumes.8
Given the subject matter a famous general devising a solution to a natural obstacle
was the most likely intermediary.
one might conjecture that the
A good picture of the historical works quarried during the compilation of the
Excerpta has been constructed from the indices auctorum preserved in the prologues
of the extant volumes (De legationibus 1, De virtutibus), or, where the prologue is
missing, identified from the contents (De legationibus 2, De insidiis, De sententiis), or,
for lost volumes, reconstructed from entries in the derivative Suda. For the present
purposes it is sufficient to observe that, for the classical and Hellenistic periods,
Constantines excerptors relied exclusively on a small number of long narrative
histories by well-known authors. Their selection must in part reflect the historical
holdings of the imperial library at Constantinople, but from a practical perspective
such works were better suited to the process of excerption than short and/or
specialised monographs. The moral dimensions and edifying intent of Constantines
great encyclopaedic project also called for reliance on reputable authorities. It can be
established with confidence, therefore, that of the historical texts known to have been
excerpted in the Excerpta only four authors could have supplied material relating to
Hannibal, namely Polybius, Diodorus, Appian and Cassius Dio. Suda 438 must
belong to one of this quartet.9
In addition to the potential author, one should consider where in an account of
Hannibals campaigns the incident described could have possibly occurred. The
technical aspects of Suda 438 will be discussed below, but the context of the episode
appears to be a march through heavily wooded terrain by a Carthaginian force
equipped with elephants; the challenge is logistical a time-consuming and labourintensive operation of forest clearance and combat with the enemy does not appear
to be imminent. Since elephants were known to be impeded by dense woodland, it
may be assumed that this forested district was too extensive to be easily circumvented
and/or that the route taken by Hannibal was one of strategic or tactical necessity or
advantage.10 The name of Hannibal is freely associated with war-elephants, but in the
7 For cross-references in the extant Excerpta to the lost
cf. EL 1.14.26,
62.31; 2.379.26; EV 1.335.19; 2.116.19, 123.26; EI 222.3; ES 131.28;
: EV
1.145.18;
: EV 1.338.7, 354.4; 2.120.6; EI 33.8. There are several candi: EV 1.99.9;
dates for the other volume(s) with military content:
: EI 207.34 (if the last two are indeed different); or
: EL 2.390.3;
: EV 1.9.20;
: ES 210.15.
8 Accordingly, A. Adler (ed.), Suidae Lexicon 2.724 (marg.) marks
438 as E = Excerpta
Constantini Porphyrogenitae quae hodie non exstant (cf. 1.xix.).
9 For detailed argumentation see C. de Boor, Zu den Exzerptsammlungen des Konstantin
Porphyrogennetos, Hermes 19 (1884), 12348; id. (n. 5 [1912]), esp. 40814; T. Bttner-Wobst,
Die Anlage der historischen Encyclopdie des Konstantinos Porphyrogennetos, BZ 15 (1906),
88120, esp. 96100. For the most convenient summary of the shape and composition of the
Excerpta Constantiniana see the masterly P. Lemerle (n. 5), 2808; recently affirmed and elaborated by B. Flusin, Les Excerpta constantiniens, logique dune anti-histoire, in S. Pittia (ed.),
Fragments dHistoriens grecs autour de Denys dHalicarnasse (CEFR 298 [Rome, 2002]), 53759.
See similar arguments in T.M. Banchich, An identification in the Suda: Eunapius on the Huns,
CPh 83 (1988), 53.
10 Elephants entangled among trees: Sall. Iug. 53.3, impeditos ramis arborum. Plut. Pyrrh. 21.7
may refer to Pyrrhus elephants impeded by a wooded riverbank at the battle of Ausculum in 279

94

PHILIP RANCE

sources his deployment of these animals is limited to two relatively brief and welldocumented periods.11 First, in the opening stages of the Second Punic War, in
summer/autumn of 218 B.C., Hannibal led an army of invasion, including 37
elephants, from the RomanCarthaginian frontier on the Ebro, through southern
Gaul and across the Alps. An unspecified number survived the Alpine transit to fight
at the battle of the Trebia (late 218 B.C.), though all but one of the remaining beasts
died during the harsh winter of 218/17 B.C.12 The authorities at Carthage voted a
reinforcement of 40 elephants in 216 B.C. (Livy 23.13.7) and an unstated number
disembarked in Italy in 215 B.C. (ibid. 41.10), but their role and significance in
subsequent operations are obscure, not least because of inconsistencies in the
evidence that have been much discussed in modern scholarship.13 Second, Hannibal
deployed at least 80 elephants at the battle of Zama in 202 B.C., without success.14
Within these contexts the passage through southern Gaul and the Alps offers by far
the most likely scenario for extensive woodland impenetrable to elephants.15 Polybius
refers to various geological, human and meteorological hazards encountered by
Hannibal; trees are never singled out as an obstacle, though Polybius does clarify the
concept of a tree line for his readers:

(for the
summits of the Alps and the parts near the tops of the passes are all quite treeless and
bare owing to the snow lying there continuously both winter and summer, but the
slopes up to half-way up on both sides are wooded and tree-covered and on the whole
habitable, 3.55.9). It is likely that in the late third century B.C. the tree line extended
significantly higher up both sides of the Alps than it does today.16 Dio also alludes to
natural obstacles that impeded the indirect paths selected by Hannibal to avoid
detection, though again trees are not specified.17 Livy records an incident in the Alps
B.C.,

as Scullard (1974), 106; Charles and Rhodan (2007), 379; however, in the most recent
editions the reading
found in all codices is emended to
(ed. K. Ziegler [19712]; ed.
corr. cum add. H. Grtner [1996]), in which case the obstacle would be mud, though I can see no
compelling reason to distrust the transmitted text. See discussion of this episode in A.B.
Nederlof, Plutarchus Leven van Pyrrhus. Historische Commentaar (AmsterdamParis, 1940),
1446, and generally R.F. Gower, The tactical handling of the elephant, G&R 17 (1948), 111,
at 5; Charles and Rhodan (2007), 37980.
11 See Scullard (1948); Gowers and Scullard (1950); Scullard (1974), 15477; J. Edwards, The
irony of Hannibals elephants, Latomus 60 (2001), 9005; and now Charles and Rhodan (2007).
The Suda entry is not adduced.
12 Polyb. 3.42.11 numbers 37 elephants at the crossing of the Rhne; cf. App. Hann. 1.4; Eutr.
Brev. 3.8.2. On the Trebia see Charles and Rhodan (2007), 3726. Sole survivor of winter 218/17
B.C.: Livy 21.56.6, 58.11; 22.2.10; Zonar. 8.24 (Dindorf 2.242.67).
13 For the meagre references to post-215 B.C. elephants and bibliography see n. 21 below.
14 Polyb. 15.11.1; Livy 30.32.4, 35.3; App. Pun. 7.40; Front. Str. 2.3.16.
15 For the countless attempts to match ancient accounts to modern topography see selectively
Walbank, Commentary, 1.3827 with older literature; D. Proctor, Hannibals March in History
(Oxford, 1971); J. Seibert, Forschungen zu Hannibal (Darmstadt, 1993), 195200; J. Prevas,
Hannibal Crosses the Alps: The Invasion of Italy and the Punic Wars (Cambridge, MA, 1998).
16 See the brief summary of historical and ecological data by G. de Beer, Alps and Elephants:
Hannibals March (London, 1955), 1447; with additional thoughts by D. Proctor (n. 15), 2057.
17 Cassius Dio apud Zonar. 8.23 (Dindorf 2.23940),
(Then Hannibal, in haste to set out for Italy, but suspicious of the more direct roads, turned aside from them and followed another, in which he met
with grievous hardships ); there follows a literary topos of icy mountain fastnesses of the sort
Polybius (3.47.612) had already criticized.

AN UNIDENTIFIED FRAGMENT OF DIODORUS

95

not found in Polybius account, when, in order to make a massive bonfire to shatter an
obstructive boulder, Hannibals troops arboribus circa immanibus deiectis detruncatisque struem ingentem lignorum faciunt (felled some huge trees that grew near at
hand, and lopping off their branches, made an enormous pile of logs).18 More
generally, the essentially pro-Barcid accounts that underlie the historical tradition of
the march from the Ebro to Cisalpine Gaul exalt the courage and endurance of
Hannibal and his army, and the crossing of the Alps in particular became a crucial
ingredient in Hannibals development as a heroic and, later, romantic figure.19 Stories
of how the Carthaginians triumphed over natural obstacles unsurprisingly accumulate in this phase of the war, including the well-documented crossing of the Rhne,
of which there were conflicting ancient accounts, and other remarkable episodes of
route-clearance and road building.20 It is not impossible that the Suda entry refers
to an otherwise unreported incident after 215 B.C., when Hannibal received new
elephants from Africa, and the Roman pursuit of Fabian tactics dictated the
avoidance of open ground in favour of hilly or obstructed terrain, though one
struggles to identify a trace of any corresponding operation or locale in the sources
for this later period. Furthermore, references to elephants after 215 B.C. occur almost
exclusively in Livys account. Modern scholarship has tended to dismiss these Livian
beasts as the exaggeration or invention of a Roman annalistic tradition, but, for the
present purposes, their disputed historicity is of less significance than the fact that
Livy clearly drew this information from a source that was not exploited by our four
Greek historians, for whom elephants are predominantly an exotic logistical challenge
of the Ebro-to-Italy march of 218 B.C.21
18 Livy 21.37.2; cf. also App. Hann. 1.4; Amm. Marc. 15.10.11; Syrianus Magister [Anon.
Byz.], De Re Strategica 18 (61.516), ed. G.T. Dennis, Three Byzantine Military Treatises ([CFHB
Series Washingtoniensis 25] Washington, D.C., 1985) (as the Anonymous Byzantine Treatise on
Strategy). On this incident see E.T. Sage, A chemical interpretation of Livy, Classical Weekly 16
(10) (19223), 736; de Beer (n. 16), 11213; J. Prevas (n. 15), 1489. For the scientific
phenomenon in antiquity see also Vitr. 8.3.19; Pliny, NH 23.27.57; Vita S. Theodori Syceotae 55,
ed. and French trans. A.-J. Festugire, Vie de Thodore de Sykn ([Subsidia Hagiographica 48]
Brussels, 1970), 1.47.
19 C. Jourdain-Annequin, Limage de la montagne ou la geographie lepreuve du mythe et de
lhistoire: lexemple de la traverse des Alpes par Hannibal, DHA 25.1 (1999), 10127; D. Hoyos,
Hannibals Dynasty: Power and Politics in the Western Mediterranean, 247183 BC (London,
2003), 111.
20 Cf. Polyb. 3.54.555.9; Livy 21.36.137.6; Nepos, Hann. 3.4. For logistical difficulties: J.F.
Shean, Hannibals mules: the logistical limitations of Hannibals army and the battle of Cannae,
216 BC Historia 45.2 (1996), 15987, at 17780. On the Rhne crossing there are as many modern
views as authors; for older bibliography see J. Seibert (n. 15), 1945; more recently: R.S.
OBryhim, Hannibals elephants and the crossing of the Rhne, CQ 41 (1991), 1215;
P. Leveau, Le franchissement du Rhne par Hannibal: le chenal et la navigation fluviale la fin
de lge du Fer, RA (2003 [1]), 2550; O.J. Schrier, Hannibal, the Rhone and the Island: some
philological and metrological notes, Mnemosyne 59.4 (2006), 50124.
21 For post-215 B.C. elephants in Livy cf. 23.18.6, 41.10, 43.6, 46.4 (cf. Plut. Marc. 12.3); 26.5.3,
11, 6.12, 912; 27.42.7. For discussion of Livys sources and reliability see Scullard (1974),
1624; Charles and Rhodan (2007), 3778, citing earlier Quellenforschungen. The accounts of
Polybius and Cassius Dio (apud Zonaras) contain no references to elephants after 217 B.C. The
only post-215 B.C. occurrence of elephants in Appian is his report of a small number used in a
nocturnal assault on the Roman siege-lines around Capua in 211 B.C. (Hann. 412, cf. Livy 26.5.3.
11, 6.12. 912), but the details of this episode are not at all consistent with Suda 438.
Diodorus account of the Hannibalic War (Books 267) is lost but for disjointed fragments; one
might, however, speculate that, given the adherence of Diodorus and Appian to a common
source (nn. 678 below), the absence of post-215 B.C. elephants from Appian (other than at
Capua in 211 B.C.) may reflect what the lost books of Diodorus contained or omitted.

96

PHILIP RANCE

Finally, in its haste to deny Polybian authorship of Suda 438 scholarship has
overlooked certain valuable technical details which are of intrinsic interest but also
point to an origin in a well-informed contemporary source. The author specifies
that Hannibal managed to traverse this wooded terrain
, by carrying the turrets of the elephants and making use of the beasts
litters to cut away the branches to the greatest height. The author clearly refers to two
and
different items of elephant equipment
. In order to facilitate the elephants passage through the trees, their
had to be removed and carried (
), presumably by soldiers or servants,
were left on the elephants, apparently as a means of cutting back the
while the
indicates that this word
branches to a maximum height. The lemma
provided the chief lexical interest for the compiler of the Suda. The basic definition of
relates to a protective barrier or parapet, notionally the height of a mans
chest, hence breastwork is the closest English equivalent, though the word is used in
connection with analogous breast-high structures, including walls, screens, battleand
ments, a ships gunwales or a crows nest. The orthographic variants
are well attested in documentary and literary sources, including Polybius
is found with the specific meaning of a turret
and Diodorus.22 In addition,
or parapet mounted on a war-elephant, obviously the context here, though this usage
is rare and, as will be demonstrated below, of considerable assistance in identifying
is otherwise undocuthe provenance of this passage.23 In contrast, the word
mented in this context; a literal derivation would suggest a small chamber, cabin or
compartment. It is difficult to escape the conclusion that in Suda 438
designates only the upper section of the apparatus mounted on each elephants back
the breastworks, screens or parapets which in the constrained circumstances of
this episode are disassembled and removed to reduce the overall height of the animal
is recognized in some
and its equipment. This more restricted sense of
left on the elephants must then be a lower section of the same
lexica.24 The
22 Polyb. 8.4.4:
, railed in and
covered by a high breastwork; Diod. 14.51.2:
, they raised up
men in crows-nests; 27.44.4:
[F
]
, the men
standing behind the screens. The forms
and
are corroborated by Suda
436:
:
; 437:
:
(= Phot. Lex.
300 [ed. C. Theodoridis]; > % (  149 [ed. I.C. Cunningham]).
(= Suda 558; > Jos. BJ 5.7.4[318]). For
as parapet, wall or barrier: Menodot. Sam. FHG 3.104 F 1 (apud Athen. Deipnos.
15.672.d); Jos. BJ 5.7.4(318); Ath. Mech. 18.11, 23.9 (with remarks of D. Whitehead and P.H.
Blyth, Athenaeus Mechanicus, On Machines ( 0), trans. and comm. [Stuttgart,
2004], 117, 125). For
as breastwork, parapet or screen: Aesch. Sept. 32; IG22.463.86;
IGRom. 4.293ai39 (Pergam. second century B.C.), 1465, 1474 (Smyrna). A breast-high part of
surface wall: P.Cair.Zen. 445 (third century B.C.). The gunwale of a trireme: IG22.1604.31. The
crows-nest on a ship: Eratosthenes fr. 60 (apud Schol. in Apoll. Rhod. Arg. 1.566), ed. K.
Strecker, De Lycophrone, Euphronio, Eratosthene comicorum interpretibus (Greifswald, 1884),
423; Asclepiades Myrleanus FHG 3.298 (apud Athen. Deipnos. 11.475a = Eustathius Thessal.
Comm. ad Homeri Odysseam [Leipzig 18256] 1.67.13); Athen. Deipnos. 5.43.e; Pollux, Onom.
1.91 (with E. Bethe [ed.] 1.29, app. crit.).
23 Diod. 2.17.8; 17.88.6 (app. crit.); Ael. NA 13.9 (Megasthenes F 36a Mller).
24 E.g. LSJ9, 813, s.v.
, also, the tower on the back of elephants, or rather the upper
part thereof, citing the current passage (as [Plb.] fr. 162b), though blurring the distinction at 1203,
s.v.
, tower on the back of an elephant; similarly A. Mauersberger (n. 3), Bd. 1.3, 1164,
s.v.
, Turm auf d. Rcken v. Elefanten oder dessen Oberteil [fr. 162b], but again Bd.

AN UNIDENTIFIED FRAGMENT OF DIODORUS

97

structure, seemingly a kind of litter or howdah, which, unlike the

, did not
substantially increase the height or breadth of the beast and thus the quantity of the
vegetation to be cut away; indeed the instrumental dative
strongly implies that the
, whatever its design, far from being an additional
impediment, provided a platform for this very task (rather than employing ladders
and ropes).25
To my knowledge, no other source makes this terminological distinction or
supplies comparable explicit data concerning a bipartite construction of elephant
turrets. It is interesting to note, however, that Scullard has already hypothesized that
such towers would presumably be made of light wooden frameworks, fenced
around with light materials, such as wicker-work. The testimony of the limited
representational evidence is ambiguous, and in some instances certainly marred by
artistic licence or guesswork, though Scullard observes that close examination of the
turreted elephant depicted on the well-known plate from Capena (Museo di Villa
Giulia, inv. 23949) suggests a number of parallel lines running down from top to
bottom, which could indicate some form of light slatting.26 Also of potential relevance is an unprovenanced engraved sardonyx (Cabinet des Mdailles, Bibliothque
Nationale, inv. 1911) depicting an (African?) elephant mounted with a low-walled
without
?), in which stand two
open platform or howdah (an
spearmen fully exposed above the hip.27 I plan a fuller treatment of this subject in a
separate study, but I stress here that it is overwhelmingly unlikely that the description
preserved in Suda 438 is the fancy of a later author writing at a time when elephant
warfare had abated, and these details suggest that this episode was first reported by a
well-informed source, who based his account on autopsy or eye-witness testimony,
and who was acquainted with technical aspects of elephantry. One might also infer
that this record of Carthaginian logistical difficulties in dense woodland during a
route-march reflects a Carthaginian standpoint, rather than Roman intelligence of
enemy activity, and probably originates from a Carthaginian witness, a consideration
to which I shall return below.
We therefore have four candidates for authorship, a likely historical setting and a
potential authorial perspective. Casaubons ascription of the fragment to Polybius
was not merely wishful thinking. The Suda contains copious entries drawn from the
Constantinian Excerpta of Polybius, who must be considered among the compilers
favourite authors, and the Suda is therefore an important editorial resource for
reconstituting the lost books of Polybius. Difficulties arise, however, when one
attempts to place 438 within Polybius narrative: his account of the passage from the
1.4, 1722, s.v.
, Turm auf dem Rcken von Elefanten. See already J. Schweighuser (n. 2
[1792]), 5.60, Possis suspicari, [
] esse pinnas vel propugnacula, quae superne cingebant
turriculas, dorsis belluarum impositas.
25 The phrase
is construed here as instrumental dative, by means of
the beasts litters, rather than pure dative, cutting away the branches to the greatest height for
the beasts litters. This interpretation is the more likely, as it would seem strange, having gone to
the trouble of removing the turrets, to leave the litters on the elephants if they added significantly
to the overall height; rather, the
are mentioned because they have a functional significance in this operation. In any case, the bipartite construction of this equipment is clear.
26 Scullard (1974), 244, with pl. 7a, and bibliography at 113, n. 7; see also J.D. Beazley,
Etruscan Vase-Painting (Oxford, 1947), 21112 with pl. 39.1; Gowers and Scullard (1950), 273,
n. 9; Goukowsky (1972), 4901, fig. 8.
27 A. Chabouillet, Catalogue gnral et raisonn des cames et des pierres graves de la
Bibliothque Impriale (Paris, 1858), 254, no. 1911; Goukowsky (1972), 4901, fig. 10; Scullard
(1974), 245, fig. 23.

98

PHILIP RANCE

Rhne to Gallia Cisalpina is extant and complete, and the fragment finds no parallels
therein (3.4755, cf. also Livy 21.308). Casaubon was not alone in assigning unidentified anonymous Suda entries to Polybius which later turned out to belong to another
author.28 Of the other candidates, Appian may be safely ruled out since his brief
account of these events is also preserved in full (Hann. 1.4). The relevant book of
Cassius Dio (14) survives as a reasonably detailed epitome in Zonaras and again there
is nothing of this incident.29 To summarize: Polybius and Appian must be excluded. It
is not impossible that our fragment derives from a fuller version of Dios work, but
there is no positive evidence to recommend such a hypothesis, and arguments to be
presented below point strongly to the fourth candidate, Diodorus.
Significantly fewer citations from Diodorus than Polybius have been identified in
the Suda. This is presumably because Diodorus Bibliotheca was relatively lacking in
lexical interest, the primary criteria for inclusion, but his rather colourless style also
hinders recognition of unattributed entries.30 The Suda contains 64 entries derived
from Diodorus, with another dozen or so suspected fragments.31 Ascription of Suda
438 to Diodorus has much to recommend it on a priori grounds. In contrast to the
narratives of Polybius, Appian and Cassius Dio, the section of Diodorus work
dealing with the Hannibalic War (Books 267) survives only as insubstantial textual
wreckage salvaged from the Excerpta Constantiniana and Excerpta Hoescheliana,
which preserve a few widely scattered fragments, mostly short and deprived of
context, from which a coherent picture cannot be reconstructed.32 It was once hoped
that an outline of these books might be dimly discerned in a skeletal rsum found in
the twelfth-century historical verses of John Tzetzes, who claims Diodorus as one of
his sources, but this farrago of clichs and errors cannot be claimed as a fragment
in any real sense.33 Diodorus account of the early phase of the war, including
Hannibals passage into Italy and the battles of 218216 B.C., has entirely perished. Of
the four candidates, therefore, only Diodorus provides a suitable lacuna into which
our fragment might be inserted.
Furthermore, insofar as the brevity of the Suda entry permits analysis of vocabulary and style, Diodorus authorship is supported by comparative lexical evidence
28 For citations of Polybius in the Suda see De Boor (n. 5 [1912]), 3957; id. (n. 5 [191419]),
247, 6581; J. Becker, De Suidae Excerptis Historicis (Bonn, 1915), 6771. For another example
of mistaken identity: H. Valesius (n. 5), 212, assigned Suda 2062 (Adler 1.186.12) to Polybius
(now [fr. 106]), an ascription retained for two centuries until T. Gaisford (ed.), Suidae Lexicon
(Oxford, 1834) noticed that the entry is in fact Diod. 14.10.3.
29 Cassius Dio apud Zonar. 8.23 (Dindorf 2.23840); the only reference to elephants in this
period is the crossing of the Rhne (239.1819).
30 For citations of Diodorus in the Suda see J. Becker (n. 28), 4755; De Boor (n. 5 [191419]),
901; A. Alder, Suidas (Lexikograph), RE 4A.1 (1932), 675717, at 702; Chamoux and Bertrac
(1993), cxxxviiicxl. The lower number of citations from Diodorus, in any case, is not accounted
for by the amount of material available to the compiler of the Suda: in the extant Excerpta
Constantiniana Diodorus is represented by 949 excerpts drawn from all 40 books, see Chamoux
and Bertrac (1993), cxxxivcxxxvii.
31 See Appendix
32 Goukowsky, Fragments II, 166 estimates that Book 26 covered the decade from the crossing
of the Alps to the battle of the Metaurus (218207 B.C.)
33 Diod. 25.19 (Walton) = J. Tzetzes, Historiarum Variarum Chiliades, ed. P.A.M. Leone
(Naples, 1968), 1.703808 (Hist. 27), citing (7034) Diodorus, Cassius Dio and Dionysius of
Halicarnassus. The Alpine crossing occupies just two lines (74950), with no mention of
elephants. This summary appears to be Tzetzes own construct, written from memory rather than
quotation, and with major blunders (e.g. Cannae follows Metaurus). Goukowsky, Fragments II,
1401 excludes most of the text.

AN UNIDENTIFIED FRAGMENT OF DIODORUS

99

from other parts of his Bibliotheca. First, the non-technical vocabulary of the fragment keenly favours Diodorus, certainly over Polybius. This is best illustrated with the
. The word
,
final clause
here meaning passage or march, is found nowhere in Polybius, once in Appians
Hannibalica and once in Cassius Dio. In contrast,
is a standard element of
Diodorus vocabulary, occurring 51 times.34 More conclusively, among these passages
of Diodorus is one instance of a
similarly made
(4.19.4) and
another of
(12.12.3).35 These coincidences of vocabulary do
not occur in the other three historians, or indeed in any classical author other than
Diodorus. Second, when we turn to technical vocabulary the pattern is yet more
revealing. We have already observed that it was the terminus technicus
that
attracted the interest of the compiler of the Suda, and the various definitions of this
term have been discussed. With the meaning turret or parapet on a war-elephant,
however,
is an exceptionally rare usage, both inherently and relative to
, the only term employed in this context by Polybius and other Hellenistic
authors. Indeed
is attested with this sense in only three other instances, two
of which occur in Diodorus.36 The three passages will be discussed in detail.
(1) Within an excursus on the history of Assyria and Media (2.134), Diodorus
provides an account of an Assyrian invasion of India led by Semiramis
(2.16.419.10), an entirely fictional episode the historicity of which was doubted even
in antiquity (cf. Strabo 15.1.56). Diodorus relates the response of the Indian ruler
Stabrobates when informed of the enemys preparations:
(8)
(having conducted a hunt
for wild elephants and multiplied by many times the number already at his disposal,
he fitted them all out splendidly with such things as would strike terror in war; and the
consequence was that during their attack, on account of their large number and the
fact that they had been furnished with turrets, they appeared like something beyond
the power of human nature to withstand, 2.17.78).
Diodorus chief source for this tale was Ctesias of Cnidus (Lenfant F 1 = FGrH
688 F 1), whom he cites eleven times, and whose Persica (written shortly after 398/7
B.C.) he almost certainly knew directly; or at least no convincing case has been made
for a Zwischenquelle.37 Semiramis theatrical attempt to counter the Indian rulers
34 Diod. 2.6.6, 54.2, 7; 3.17.3, 72.4; 4.19.4, 22.5, 23.1; 5.29.1; 11.56.4, 74.2, 75.2, 80.1; 12.12.3;
13.89.3, 112.5; 14.19.9, 26.4, 5, 27.2, 28.2, 47.5; 15.53.2; 16.17.5, 31.3, 42.1, 84.5; 17.4.4, 24.2,
32.3, 53.3, 65.2, 94.2, 106.1; 18.59.3; 19.5.3, 19.1, 21.2, 24.5, 26.5, 31.2, 38.5, 80.2; 20.29.7, 42.2;
22.10.1; 29.23.1; 33.14.4; 34/5.2.28, 4.1, 8.1. Cf. App. Hann. 40; Cassius Dio 56.20.3.
35 Diod. 4.19.4:
; 12.12.3:
.
36 Goukowsky (1972) passim fails to appreciate the rarity of this usage of
and
erroneously employs it as if it were widely used vocabulary; similarly Scullard (1974), 2402;
Charles (2008).
37 Ctesias F1b (1619). Fragments of Ctesias are cited after D. Lenfant (ed., French trans.
and comm.), Ctsias de Cnide, La Perse, lInde, autres fragments ([Bud] Paris, 2004). Date of the
Persica: F. Jacoby, Ktesias1, RE 1.9 (1922), 203273, at 20346; Lenfant (ed.), xviii, xxiixxv.
For Diodorus use of Ctesias see Bigwood (1980), citing earlier bibliography. Diodorus direct
knowledge of the Persica was demonstrated at length by P. Krumbholz, Diodors assyrische
Geschichte, RhM 41 (1886), 32141; id., Zu den Assyriaka des Ktesias, 50 (1895), 20540; 52
(1897), 23785; endorsed and supplemented by Bigwood (1980), 1969.

100

PHILIP RANCE

elephants is a central theme that runs through the story of her expedition, and there is
no doubt that the elephants themselves are an original element of Ctesias narrative.38
There are very good reasons, however, for thinking that Diodorus did not find the
reference to turrets (
) in Ctesias work but inserted this detail himself.
Diodorus substantially altered the style and content of his source, and supplemented
the Persica with his own remarks as well as material drawn from other authorities.39
Although he occasionally allows his model to influence his language, Diodorus
narrative is for the most part framed in his own choice of vocabulary and little can be
reconstructed of Ctesias original wording. This is certainly the case with the current
passage, in which a stereotypical battle account has been fashioned from words and
phrases favoured by Diodorus elsewhere in his work.40 Furthermore, Diodorus
interpolations are sometimes betrayed by details that must post-date Ctesias era or
conflict with the extant fragments of his works. The most glaring anachronisms
involve the unhistorical retrojection of military technology: hence the Assyrians have
at their disposal stone-throwing artillery (
) which was first developed in
the Greek world in the mid fourth century B.C., and was thus unknown even in Ctesias
day.41 In this context, it is widely accepted that throughout the Indian subcontinent in
38 Diod. 2.16.817.8, 18.66. Cf. Suda
220 (Adler 4.339) for another account of Semiramis
and the elephants, which must also derive from Ctesias (though omitted by Lenfant ed. [2004]),
on which see P. Krumbholz (n. 37 [1897]), 27980. There are three other references to elephants in
Ctesias Persica and Indica: (1) autopsy of an elephant with its Indian trainer at Babylon: Ctesias
F 45b = Ael. NA 17.29, cf. 5.55 (cf. Aristotle, HA 9.1.610a224); see J.M. Bigwood, Ctesias
description of Babylon, AJAH 3 (1978), 3252, at 323; id., Aristotle and the elephant again,
AJPh 114.4 (1993), 53755, at 5423. (2) Military applications of elephants in India: Ctesias F
45b = Ael. NA 17.29; Ctesias F 45.7 = Photius, Bibl. 72.45a (cf. Aristotle, HA 9.1.610a1522),
with Lenfant (ed.), 187, 31922, n. 890; Goukowsky (1972), 4745; K. Karttunen, India in Early
Greek Literature ([Studia Orientalia 65] Helsinki, 1989), 63, n. 460; id., India and the Hellenistic
World ([Studia Orientalia 83] Helsinki, 1997), 188. (3) Cyrus encounters elephants during a
campaign against the Derbices: Ctesias F 9.7 = Photius, Bibl. 72.36b.
39 Bigwood (1980), 199200. The most detailed study of Diodorus style is J. Palm, ber
Sprache und Stil des Diodoros von Sizilien. Ein Beitrag zur Beleuchtung der hellenistischen Prosa
(Lund, 1955), though his demonstrations do not include the section examined here. For
Diodorus other named sources: Cleitarchus cited at Diod. 2.7.3 (= FGrH 137 F 10); an unknown
Athenaeus at Diod. 2.20.35; see J.M. Bigwood (n. 38 [1978]), 45, n. 11; id. (1980), 199, n. 20,
2023, 2056; J. Boncquet, Diodorus Siculus (II,134) over Mesopotami: Een historische kommentaar ([Verhandelingen van de koninklijke academie voor wetenschappen, letteren en schone
kunsten van Belgi. Klasse der letteren 122] Brussels, 1987), 757.
40
For Diodorus stereotyped battle-pieces see C. Vial (ed.), Diodore Livre XV ([Bud] Paris,
1977), xxxii. The account of the battle between Semiramis and Stabrobates features vocabulary
typical of Diodorus, especially
and its cognates (2.16.3, 8; 2.17.7, 19.4). See
J. Palm (n. 39), 167; Bigwood (1980), 201. G. Goossens, LHistoire dAssyrie de Ctsias, AntCl 9
(1940), 2545, at 412 discerns in the engagement between Semiramis and Stabrobates (2.19) the
dynamics of a Hellenistic battle.
41
Diod. 2.27.1. See G. Goossens (n. 40), 43; Bigwood (1980), 2045; J. Boncquet (n. 39),
1645. For Diodorus interest in the technical aspects of warfare see Chamoux and Bertrac
(1993), liliv; for artillery in particular see now H.M. Schellenberg, Diodor von Sizilien 14,42,1
und die Erfindung der Artillerie im Mittelmeerraum, Frankfurter elektronische Rundschau zur
Altertumskunde 3 (2006), 1423 (http://www.fera-journal.eu). Similarly, Diod. 2.16.4 comments
that the Indian elephant is larger than the African, a fact of which Ctesias was almost certainly
ignorant, see J.S. Romm, Aristotles elephant and the myth of Alexanders scientific patronage,
AJP 110.4 (1989), 56675, at 5745; J.M. Bigwood (n. 38 [1993]), 543, n. 31, 550, n. 66; see
further discussion see n. 60 below. Diodorus account of Semiramis Indian campaign refers to
post-Ctesias events: 2.5.6 (Dionysius of Syracuse), 5.7 (Punic Wars), 17.3 (Perseus of Macedon).
See also anachronistic geographical terminology reflecting Seleucid rather than Ctesian (i.e.
Achaemenid) usage: Bigwood (1980), 2001. It has also been observed that Diodorus treatment

AN UNIDENTIFIED FRAGMENT OF DIODORUS

101

the late fifth / early fourth century B.C., and probably much later still, war-elephants
were not furnished with turrets or howdahs and so, even granting that Semiramis
expedition is an unhistorical fantasy, this equipment simply could not have been
mentioned by Ctesias or his Persian informants. Accordingly, all modern commentators concur that the
on Stabrobates elephants are Diodorus own
invention inspired by Hellenistic practices.42
(2) In an account of Alexanders victory on the River Hydapses in 326 B.C., Diodorus
describes how the rajah Porus was heavily wounded and attempted to withdraw his
elephant from the fighting (17.88.6). The text of Diodorus 17 depends on two
manuscript prototypes which at this point preserve different readings. Parisinus gr.
1666 (= R) reads:
, having lost much
blood on account of the multitude of his wounds, he fainted and, slumping over the
parapet, sank to the ground. Directly above
occurs the word
written
in the same hand. It cannot be determined whether it was the copyist of R himself
who first inserted this supralinear variant or whether he mechanically reproduced
what he found in his exemplar, though in either case its presence surely expresses a
Byzantine editors doubt or incomprehension of the word
in the main
text.43 In Mediceo-Laurentianus gr. 7012 (= F), however, the sole family II manuscript to contain Diodorus 17, one reads only
. Editorial choice of variant
therefore determines whether the enfeebled Porus fainted and tumbled over the
parapet on his elephant or, in the absence of a turret, merely slumped over the beast
itself. The reading
was preferred by Henri Estienne in the editio princeps (1559)
and retained by all subsequent editors, including the influential edition of Peter
Wesseling (1745), even though he stressed in his commentary the superior merits of
.44 Goukowsky has especially pressed the case for
, arguing that a
turret is not mentioned by the other authors who, like Diodorus, drew on the so-called
Vulgate or Cleitarchan tradition for this episode.45 Moreover, Goukowsky conof Assyrian and Median history (2.134) contains psychological characterization of individuals
consistent with moralizing stereotypes found elsewhere in his Bibliotheca, which cannot therefore
be authentic elements of Ctesias Persica, see Bigwood (1980), 202.
42 Goukowsky (1972), 475 with n. 10; Scullard (1974), 356, 241; J.M. Bigwood (n. 38 [1993]),
243, n. 31; Lenfant (ed.), 45, n. 229 (F1b 17.8). For the howdah in ancient India see n. 55 below.
43 Goukowsky (1972), 473 asserts that in R the reading
est indiscutablement une
variante de premier main, que le plus proche parent connu de R, le Vaticanus 132, donne lui aussi
selon la mme disposition. Vaticanus gr. 132 (= Y) cannot be used to illuminate the textual
history of this variant in R, however, because Y is a fifteenth-century apograph of R and not an
independent witness, as Goukowsky seems to believe here and reiterates, more guardedly, at
Goukowsky, Diodore XVII, xlix. For the orthodox view of these codices see Chamoux and
Bertrac (1993), ciicv, cixcx, with stemma codicum at cxxi.
44 See most recently Goukowsky, Diodore XVII, 1234. See P. Wesselingius (ed.), Diodori Siculi
Bibliothecae Historicae Libri Qui Supersunt (Amsterdam, 1745), 2.229.1819, Quis controversiam his movisset, nisi veteres membranae ansam praeberent?
, quod ostentant, non
plane incongruum est Itaque id si reliquit Auctor, indicabitur Porum ex ea machina in
terram versum corruisse. Nihil tamen decerno, quod & vulgatum tueri se possit.
45 Goukowsky (1972), 4734. Cf. Curt. 8.14.31. The Vulgate tradition did, however, envisage
Porus elephants themselves as towers within the metaphorical rampart formed by the
Indian infantry: Diod. 17.87.5:
; Curt. 8.14.13: Beluae dispositae inter armatos speciem turrium procul
fecerant; Polyaenus, Strat. 4.3.22:

102

PHILIP RANCE

vincingly demonstrates, albeit by an argumentum e silentio, that war-elephants were

not equipped with turrets or howdahs of any kind in the Punjab or (before c. 300280
B.C.) in the armies of the Diadochi, who initially obtained their elephants and
elephant-handling techniques from north-western India. Rather, elephants in this
region and period were manned either by a mahout alone or assisted by a single spearor sarissa-armed rider (
) sitting astride the animals back.46 Goukowsky
concludes that if Diodorus did write
it could only have been a slip of his
pen, a prospect he deems unlikely, and so he prefers to discern in this reading une
correction due un grammairien trop rudit.
Goukowskys remarks on the Vulgate tradition and the contemporary utilisation of
war-elephants are not disputed here, but their relevance and value in resolving this
particular textual crux are illusory. In particular, Goukowskys appeal to the Realien
of Indian warfare in the late fourth century B.C. is both unnecessary and irrelevant; it
is Diodorus opinion that matters. The previously cited passage of Diodorus 2.17.8
(as 1) demonstrates beyond doubt that the first-century B.C. historian believed,
however inaccurately or anachronistically, that ancient Indian war-elephants were
furnished with turrets; that he was disposed to interpolate this technical detail into his
source material, and that the very word he used to denominate this item of equipment
was the exceptionally rare term
.47 One may also legitimately doubt whether
even the most erudite and meticulous of Byzantine scribes, upon encountering
in the current passage, would have found reason to distrust this unexceptionable
, which can
reading and conjectured instead the recherch classical usage
only be considered the lectio difficilior and thus the less likely of the two readings to
be Byzantine scribal surmise. Goukowskys explanation is in any case incompatible
with the codicological evidence: the configuration of the variant readings
and
in R leaves no doubt that, whoever was responsible, the supralinear
must have been the later insertion into an exemplar that originally read only
, and not vice versa.48 Furthermore, upon closer inspection the competing
claims of codices R and F prove to be no contest at all: the mid-tenth-century R is
more often a more reliable witness than the fifteenth-century F; a perusal of
Goukowskys apparatus for Diodorus 17.88 reveals that of the eighteen other
divergent readings in R and F in this chapter, R is correct in seventeen instances, the
sole exception being a minor orthographic variant.49 Indeed, it may even be doubted
; it is true that in F the text
whether F is an independent witness to the reading
(cf. Livy 28.14.4 for a similar metaphorical usage). For the vexed question of the Vulgate tradition see selectively J.R. Hamilton,
Cleitarchus and Diodorus 17, in K.H. Kinzl (ed), Greece and the Eastern Mediterranean in
Ancient History and Prehistory. Studies presented to Fritz Schachermeyr on the Occasion of his
Eightieth Birthday (BerlinNew York, 1977), 12646; N.G.L. Hammond, Three Historians of
Alexander the Great: The So-Called Vulgate Authors, Diodorus, Justin, and Curtius (Cambridge,
1983); L. Prandi, Fortuna realt dell opera di Clitarco ([Historia Einzelschriften 104] Stuttgart,
1996).
46 Goukowsky (1972), 47492, 4978; summarized by Scullard (1974), 2402.
47 Aside from the word
, these two passages of Diodorus share typically Diodoran
descriptive vocabulary: 2.16.2:
;
2.17.7:
(
)
;
17.87.4:
.
48 Note also that the only other instance of a supralinear correction in this section of R is
manifestly wrong and conflicts with the common reading in both R and F: 17.89.5
RF :
Rsl.
49 Goukowsky, Diodore XVII, 1224 app. crit. The exception is 17.88.4
F:
- R.

103

AN UNIDENTIFIED FRAGMENT OF DIODORUS

of Diodorus 1720 derives independently from a majuscule archetype

and thus
embodies an authentic tradition, but the text has been modified by a late Byzantine
editor in a process of recension that involved, inter alia, introducing corrections from
other manuscripts, including R. The editor of F would therefore have been aware of
the two variants
and
in R, and the presence of
in F is just as
likely to be a product of his editing as a genuine tradition from .50 Accordingly, the
stylistic, lexical and codicological evidence, individually and cumulatively, overwhelmingly favours
against
, regardless of whatever wider historical
thesis Goukowsky seeks to present.
(3) The third and only non-Diodoran instance occurs in a fragment of the Indica of
Megasthenes preserved in Aelians De natura animalium 13.9:
{**}

, The chariot also carries two crewmen. The war-elephant, on

what is called the turret, or, by Jove, on his bare and unencumbered back, carries
three warriors, {of whom two} shoot to either side, and the third at the rear, while a
fourth man carries in his hand the goad with which he controls the beast, much in the
same way as a pilot and ships captain control a ship.51 The same passage of
Megasthenes underlies a much more abbreviated report in Strabos Geographica:
, There are two crewmen in the chariot
in addition to the driver, while the driver of the elephant is one of four, with three
others shooting bows from it.52 Again the text is problematic. In his Indica Megasthenes reported his observations during an embassy to Patliputra, capital of the
Mauryan Empire, during the reign of Chandragupta, a journey conventionally placed
c. 304/3 B.C., but conceivably dating to any point c. 319/18288 B.C.53 The fuller text in
Aelian states that a Mauryan elephant crew normally comprised three archers plus a
mahout, and that some sort of turret or howdah (
) was one option in
combat, alternative to all four men perching on the elephants back and shoulders.
Goukowsky influentially dismissed the reference to a
as Aelians interpolation, noting that this detail is not found in Strabos version, and again stressing
the otherwise heavy silence of both literary and representational sources with regard
50
See F. Bizire (ed. and French trans.), Diodore de Sicile, Livre XIX ([Bud] Paris, 1975),
xxxxii; Chamoux and Bertrac (1993) ciiicv, cxxi, cxxxvicxxxvii, who characterize F as rsultat
dun travail philologique tardif (civ).
51 Aelian, NA 13.9, ed. R. Hercher (Leipzig, 1864; repr. Graz, 1971) 323.1320 (with app. crit.
at xlv) = E.A. Schwanbeck (ed.), Megasthenis Indica (Bonn, 1846; repr. Amsterdam, 1966), F
35.4 = Mller FHG 2.431 F 36a. It is mystifying why Jacoby FGrH 715 omits this fragment, yet
allows Strabo 15.1.52 = F 31 (see following note). Jacobys inconsistency caused unjustified
doubts concerning uncertainty of origin in Scullard (1974), 241.
52 Strabo 15.1.52 = Megasthenes: Schwanbeck F 34.15 = Mller FHG 2.431 F 36a = Jacoby,
FGrH 715 F 31 (3 C.2 634.1214).
53 See A.B. Bosworth, The historical setting of Megasthenes Indica, CPh 91.2 (1996),
11327, who remarks on the fragility of the traditional date of 304/3 B.C. and argues for earlier
contact, though, in my view, the bureaucratic system Megasthenes describes is more suggestive of
the Mauryan Empire in its well-established phase. The terminus ante quem is the death of
Chandragupta in 288 B.C. For a summary of earlier scholarship see A. Zambrini, Gli Indika di
Megasthene, ASNP3 12 (1982), 71149; 15 (1985), 781853; K. Karttunen (n. 38 [1997]), 7094.

104

PHILIP RANCE

to the use of howdahs in India and the Hellenistic world before c. 300280 B.C.,
around when, Goukowsky contends, elephant turrets or carriages were first devised in
the West, probably by or for Pyrrhus in Epirus (and presumably transmitted back to
India at some later date or developed there independently).54
The complex problem of the origin and diffusion of the howdah awaits definitive
discussion, which will not be attempted here, but I note that the Indian sources are
more complicated and equivocal than Goukowsky allows, and his thesis runs the risk
of homogenizing the evidence for elephant warfare across a vast and diverse subcontinent.55 Furthermore, the intertextual case for Aelians interpolation of content is
,
weak Strabos version is in every way more cursory and the absence of a
or indeed any explanation of how the crew was mounted, is just one of several
respects in which Strabos account is deficient. Scullard endorsed Goukowskys
contention that the mention of towers could well be an addition by Aelian himself on
the basis of its (sic) frequent mention by Hellenistic writers whom he knew, but, like
in this context
Goukowsky, Scullard failed to appreciate the rarity of
is the only word for an elephant turret or howdah documented in
given that
Greek sources other than Diodorus, it is difficult to divine in which Hellenistic
writers Aelian would have found this obscure usage?56
The evidence does not admit a simple solution. On the one hand, Aelian (c. A.D.
175235) was a consummate rhetorician whose stylistic interests created opportunities for interpolation. Comparison with other Megasthenes-derived sources reveals
that, while Aelian preserved some of the wording of his model, there can be no doubt
that the received text of this fragment also reflects his own idiomatic tastes.57 If not
might have arisen from Aelians known
Megasthenian, the reference to a
predilection for archaic or recherch vocabulary.58 If this be the case, it is possible that
54 Goukowsky (1972), esp. 4889. He does not consider what possible counter-current/s might
have carried this allegedly Greek innovation in elephant tactics back to the source of the
elephants. Goukowskys view is accepted by Scullard (1974), 2412.
55 Indologists have observed that other apparent discrepancies in the Greek sources for ancient
Indian elephantry may be resolved by the simple realization that customs and techniques varied
across India; in particular, the petty dynasties and aristocratic polities of the western Punjab and
Indus valley, as observed by the authors who accompanied Alexander in 3276 B.C., differed in
significant respects from the centralized bureaucracy and powerful state-funded standing army
of the Mauryan Empire, as described by Megasthenes. See T.R. Trautmann, Elephants and the
Mauryas, in S.N. Mukherjee (ed.), India. History and Thought. Essays in Honour of A.L. Basham
(Calcutta, 1982), 25481, esp. 25460. The very fact that Megasthenes states (on the combined
testimony of Aelian and Strabo) that a Mauryan war-elephant carried a four-man crew should at
least alert us to the likelihood that the military practices of the Gangetic plain differed from the
mahoutmodel Goukowsky discerns in north-western India. The Indian sources do
not permit firm conclusions concerning the early history of the howdah, see remarks and older
bibliography in D. Singh, Ancient Indian Warfare with Special Reference to the Vedic period
(Leiden, 1965), 75, 7884; S.K. Bhakari, Indian Warfare: An Appraisal of Strategy and Tactics of
War in the Early Medieval Period (New Delhi, 1981), 629; G.N. Pant, Horse and Elephant
Armour (Delhi, 1997), 10711 (to be read with caution).
56 Scullard (1974), 2412, see n. 36 above.
57 The phrase
, for example, certainly belongs to Aelian, cf. NA 1.29.4; 5.16.9;
6.53.9; 11.15.3; 12.6.7, 15.6; 14.10.3, 26.41; 17.11.14; ep. rust. 18. With equal certainty, however,
other vocabulary can be deemed the authentic wording of Megasthenes, e.g.
; see
analysis by B.C.J. Timmer, Megasthenes en de Indische Maatschappij (Amsterdam, 1930), 1567,
1647.
58 Note, for example, that Ael. NA 13.9, 22 is the only author to use the word
with the
sense of an elephant-goad or ankus. The same term may possibly be discerned in a garbled
lemma in Hesych. Lex. 399 (ed. K. Latte 1.17):

AN UNIDENTIFIED FRAGMENT OF DIODORUS

105

Aelians choice of
over
owes nothing to earlier Hellenistic histories,
as Scullard reasoned, but is a purely metaphorical allusion inspired by his comparison
with the sense of a gunwale
of the elephant to a ship and the attested use of
, the
or crows-nest.59 The epexegetical construction
so-called turret , at least implies Aelians awareness that he is dealing with arcane
in
jargon. On the other hand, proponents of the Megasthenian pedigree of
Aelian might find modest support in the occurrence of the same usage in Diodorus
Bibliotheca. Of all the historians who refer to elephant turrets Diodorus is distinctive
and he alone used Megasthenes
in two respects: he alone employs the term
Indica as a source. More specifically, Diodorus amplified his primarily Ctesias-based
account of Semiramis invasion of India (2.1617) by inserting additional Indiarelated material drawn from the same sources as he later exploited for his main
treatment of India (2.3542); these sources included Megasthenes.60 The exclusive
terminological concurrence between Aelian and Diodorus, therefore, opens the
from Megaspossibility that both authors imported this atypical usage of
thenes, though uncertainty about the nature and extent of Diodorus familiarity with
the Indica means that this line of reasoning cannot advance beyond conjecture.61

. It is pointed out by R. Goossens, Gloses indiennes dans le Lexique dHsychius, AntCl

12 (1943), 4755, at 523, that
must result from negligent copying; he suggests a
conflation of
(cf. Sanskrit anka, ankusa), though LJS9 1256, s.v.
consider
another possibility. For alternative terminology cf. Aristotle, HA 9.1.610a28:
; App. Pun. 7.43:
.
59 See n. 22 above.
60 At 2.16.3 Diodorus briefly introduces the reader to India insofar as this background is
required for narrating Semiramis expedition. This survey was not in Ctesias, but closely
resembles and anticipates Diodorus main description of India at 2.35.3, see P. Krumbholz,
Wiederholungen bei Diodor, RhM 44 (1889), 28698, at 2936; Bigwood (1980), 206;
J. Boncquet (n. 39), 11415; Lenfant (ed.), 43, n. 222 (F 1b 16.34). Similarly, the account of
Semiramis invasion plan at Diod. 2.16.4 prompts the observation that Indian elephants are
larger than African, a fact unknown to Ctesias (see n. 41 above) which Diodorus drew from the
same source as he used for his main account of India at 2.35.4 (cf. also 42.1). The verbal parallels
leave no doubt that the same source lies behind this doublet: Diod. 2.16.4:
; 2.35.4:
. Diodorus source for this statement cannot be
identified with certainty, as several authors appear to have made similar remarks. Onesicritus
expressed this view (cf. Strabo 15.1.43), though the data on elephants at Diod. 2.42 in every other
respect conflict with the fragments of Onesicritus (FGrH 134 F 14). Megasthenes remains a more
likely possibility, as Jacoby, FGrH 2B.2 474 avers, though in his collection of the fragments of the
Indica he puzzlingly admits only Diod. 2.35.4 (715 F 4) but omits the identical Diod. 2.16.4.
Jacobys inconsistency led Scullard (1974), 60 with n. 29 into some confusion.
61 The optimistic view of early scholarship (E.A. Schwanbeck [n. 51], 578, 8593) that
Diodorus 2.3542 is a straightforward epitome of Megasthenes Indica has been abandoned. This
whole section is included by Jacoby FGrH 715 as Megasthenes F 4, but printed in small type in
recognition of the lack of explicit attribution. It is generally accepted that Diodorus treatment of
India draws (directly or indirectly) on Megasthenes, but that he also took India-related material
from (an)other Hellenistic source(s) (possibly Onesicritus via Cleitarchus). The subject would
benefit from reappraisal, see P. Krumbholz (n. 60), 2936; Jacoby FGrH 3.C.2 606 app. crit.;
B.C.J. Timmer (n. 57), 1921; O. Stein, Megasthenes, RE 1.29 (1931), 230326, at 26771;
T.S. Brown, The merits and weaknesses of Megasthenes, Phoenix 11 (1957), 1224, at 1518,
212; R.C. Majumdar, The Indika of Megasthenes, JAOS 78 (1958), 2736; K.D. Sethna,
Rejoinder to R.C. Majumdar, JAOS 80 (1960), 2438; R.C. Majumdar, The Surrejoinder to
K.D. Sethna, JAOS 80 (1960). 24850; K. Karttunen (n. 38 [1997]), 73.

106

PHILIP RANCE

To summarize: the textual history of the Suda requires that 438 must derive from
one of four authors: Polybius, Appian, Cassius Dio or Diodorus. All the evidence
historical, textual, stylistic, lexical supports the view that this is a fragment of
Diodorus. His Bibliotheca offers by far the most likely lacuna; the non-technical
vocabulary of the fragment exhibits close verbal coincidences with Diodorus extant
text, and with no other work; and, of the four candidates, the rare usage of
in the sense of a turret or parapet on a elephant is unique to, and even diagnostic of,
Diodorus, and marks him out from all the Hellenistic historians who universally
. Indeed, even if the field were not
denominate this equipment using the word
limited to these four historians, Diodorus would still be the most likely suspect.
I conclude with some brief remarks concerning the potential historical significance
of Suda 438 and the possible source that ultimately underlies this anecdote about
Hannibal. It is well known that in war the Indian elephant (Elephas maximus) can be
furnished with a turret or howdah, if required, but whether this was the case with the
smaller African forest elephant (Loxodonta cyclotis) used by Ptolemaic, Carthaginian
and Numidian armies has long been controversial.62 The most recent and comprehensive review of the evidence concurs with the majority of earlier scholarship that
African forest elephants did not ordinarily carry turrets, though this view necessarily
entails an argument from silence, which is occasionally broken by awkward testimony
to the contrary that can only be discounted by special pleading.63 The positive
evidence for specifically Carthaginian use of elephant turrets is undeniably slight, but
even those studies which argue a negative case concede the possibility of the exceptional
use of turrets in a variety of contexts ceremonial, propagandistic or poliorcetic.64
62 For the different species see Scullard (1974), 1531, 603; R. Sukumar, The Living
Elephants: Evolutionary Ecology, Behavior and Conservation (Oxford, 2003), 456, 524; Charles
and Rhodan (2007), 3647.
63 See now Charles (2008), who essentially affirms the views of Scullard (1948); (1974), 2405.
Problematic for the negative case are explicit references to elephanti turriti of Juba I of Numidia
in 46 B.C. in B.Afr. 30.2, 41.2, cf. 86.1 (elephantosque LXIIII ornatos armatosque cum turribus
ornamentisque). The author of B.Afr was contemporary, well informed and conversant with
military technicalia, while his sparse Caesarian diction is free of gratuitous literary embellishment, see most recently M. Mller, Das Bellum Africum: Ein historisch-philologischer
Kommentar der Kapitel 147 (Diss. Trier, 2001), 3346, with an ample bibliography. A coin of
Juba II, dated A.D. 21/2, depicts an unambiguously African elephant bearing a turret, see
J. Mazard, Corpus Nummorum Numidiae Mauretaniaeque (Paris, 1955), 103, no. 276, with pl. at
247; Scullard (1974), pl. 23.j. Also of possible relevance here is a fragment of Juba IIs writings on
elephants (FGrH 275 F 50), which alleges that the Libyans in the past branded some
war-elephants by engraving the symbol of a tower (
) on their tusks. See also Polyb.
5.84.27, where elephants deployed by Ptolemy IV at the battle of Raphia in 217 B.C. expressly
carry turrets, though one school of thought has sought to explain away this evidence by inferring
that Ptolemy fielded a force of predominantly (unturreted) African elephants supplemented by a
small number of turreted Indian beasts, and postulating Ptolemaic access to and/or breeding of
the Indian species, see Gowers and Scullard (1950), 2727; Scullard (1974), 1434; and now
M.B. Charles, Elephants at Raphia: reinterpreting Polybius 5.845, CQ 57 (2007), 30611.
64 The testimony of Lucr. 5.13024; Sil. Pun. 4.599; 9.23940, 5778; 17.621; Juv. 12.110 can
be safely disregarded as poetic embroidery. Scullard assembles the meagre representational
evidence for certain or probable Hannibalic elephants equipped with towers: (1) small silver coins
rev. depicting turreted elephant (species uncertain) probably issued by rebel Campanian cities
post-216 B.C.; (2) a phiale from Cales in Campania; (3) a terracotta figure from Pompeii
(insecurely dated). See Gowers and Scullard (1950), 2789, pl. 16b; Scullard (1974), 1701, 1767,
243 with pls. 10.ab, 22.h. On this basis Scullard (1948), 1612 and n. 9 denies Carthaginian use
of towers, but at 166 allows exceptional circumstances; repeated at Scullard (1974), 243, with a
chronological qualification: it is safer to believe that the Carthaginians in their earlier use of
elephants did not encumber them with towers [N]evertheless some archaeological evidence

AN UNIDENTIFIED FRAGMENT OF DIODORUS

107

Once it is admitted that African forest elephants not only could but sometimes did
carry turrets, residual objections on the grounds of the perceived physical inferiority
of the species become doubtful.65 That being the case, regardless of whether the
turrets were, in the event, employed frequently or exceptionally, they would still have
to be transported when a Carthaginian army took the field, rather than improvised on
campaign, and presumably the elephants themselves were the best vehicles for this
load. It has been conjectured that after 215 B.C. Hannibal may have had at his disposal
some (or at least one) Indian elephants, allegedly supplied to Carthage by Ptolemy IV
(221205/4 B.C.), who had captured Seleucid (Indian) elephants at Raphia in 217 B.C.;
according to this view, any reference to turreted elephants after this date could relate
instead to the Indian species. The tissue-thin evidence for this hypothesis hardly
compels, however, and Romano-Egyptian friendship throughout the Second Punic
War appears at least to rule out Ptolemy IV as a supplier of precious Indian elephants
to Romes enemy.66 Consequently, acceptance of the historicity of the episode
reported in Suda 438 controverts only those previous studies which dogmatically
contend that African forest elephants were never, in any circumstances, furnished with
turrets, a position which hardly seems tenable.
It was demonstrated above that both instances of
in Diodorus (2.17.8;
17.88.6) are his own interpolation of an unhistorical detail that was not present in his
and
source. This does not appear to be the case at Suda 438, in which
are not mere descriptive embellishment but intrinsic and integral elements of
the account, and it is overwhelmingly unlikely that this technical distinction, and thus
the entire episode, is the invention of Diodorus; in short, he is the conduit not the
source. It has also been suggested that this incident derives ultimately from eyewitness testimony, most probably Carthaginian, and possibly with military expertise.
This begs the obvious question whether Diodorus could have had access to such
information. Although Diodorus account of the Second Punic War survives as a few
isolated fragments that merely hint at the content and arrangement of lost Books
267, exhaustive Quellenanalyse has established some possible sources. While detailed
interpretations differ, it has long been recognized that comparison of the remains of
Diodorus with other extant sources permits two basic conclusions: first, for this
does suggest that towers were carried on occasion during the Hannibalic War. Charles and
Rhodan (2007), 3668 similarly nuance a generally negative argument: (368) we assume that
African elephants used in warfare by Carthage were not equipped with turrets (at least in pitched
battles). More positive is Goukowsky (1972), 490, n. 67, Il semble que les Carthaginois naient
quip que tardivement leurs lphants de tours, mais il est certain quune partie au moins de
ceux dHannibal taient ainsi arms. Other opinions, both scholarly and amateur, are assembled
by Charles and Rhodan (2007), 366 with n. 19.
65 To my knowledge, this proposition originates in W. Gowers, African elephants and ancient
authors, African Affairs 47 [188] (1948), 17380, at 179; repeated with minor rewording in
Scullard (1948), 1612; (1974), 242; similarly de Beer (n. 16), 104; J.F. Lazenby, Elephants,
OCD3, 520; id., Hannibals War: A Military History of the Second Punic War (Warminster,
19982), vii, 1516.
66 The case for Hannibals Indian elephant(s) relies on an imaginative interpretation of inadequate numismatic and onomastic evidence, see Gowers and Scullard (1950), 27783; H.H.
Scullard, Ennius, Cato, and Surus, CR (N.S.) 3 (1953), 1402; Scullard (1974), 1706; F. de
Visscher, Une histoire dlphants, AntCl 29 (1960), 5160 at 545; Charles and Rhodan (2007),
366. I plan to examine this problem in a separate study. For Egypts official neutrality and actual
assistance to Rome during the Second Punic War see L.H. Neatby, Romano-Egyptian relations
during the third century BC, TAPA 81 (1950), 8998; A. Lampela, Rome and the Ptolemies of
Egypt. The Development of their Political Relations 27380 BC ([Commentationes Humanarum
Litterarum 111] Helsinki, 1998), 3386.

108

PHILIP RANCE

period Diodorus owes nothing to Polybius; second, numerous striking correspondences with Appian, and to a lesser degree with Livy, indicate that Diodorus main
source for the Second Punic War was a historical work, written in Greek, which
belonged to or drew on the Roman annalistic tradition. This author was certainly
pre-Livian, and possibly pre-Polybian, and was peculiarly well informed about
domestic politics at Carthage.67 Although opinions have differed, the most recent and
comprehensive assessment of the evidence has tentatively advanced the candidature
of Lucius Cincius Alimentus.68 With regard to Suda 438, this proposal is especially
tantalizing. The plebeian senator L. Cincius Alimentus (c. 240190 B.C.) wrote a
history of Rome in Greek that ran from the foundation of the city to the end of the
Second Punic War.69 Alimentus participated in the war against Hannibal: he commanded two legions as praetor in Sicily in 210 B.C., where he remained as propraetor
in 209/8 B.C. After his return to Rome in 208 B.C., he was a member of the three-man
mission sent to the consul T. Quinctius Crispinus in Capua. More significantly,
Alimentus reported in his work that he had been taken prisoner by Hannibal but
apparently treated with respect, and his confinement (c. 208201 B.C.) placed him as a
privileged witness to Carthaginian perspectives. During his detention he had the
opportunity to converse with Hannibal himself about his previous conduct of the
war. As revealed in a fragment preserved by Livy, Hannibal specifically provided
Alimentus with detailed information on the Carthaginian army during its crossing of
the Rhne and the Alps.70 L. Cincius Alimentus, already proposed as the chief source
for Diodorus 267, could therefore be the source for the episode in Suda 438.
Diodorus dependence on this annalistic tradition for his principal model does not
preclude his use of supplementary material from another source(s). As a second
possibility, it is interesting to note that among the fragments of Diodorus 26 is an
explicit reference to a Carthaginian account of the Second Punic War: Sosylus the
Lacedaemonian wrote a work about Hannibal in seven books. The fragment supplies
no context and the reference does not in itself prove that Diodorus had read this
work, though it does not seem too incautious to assume that he had a good reason to
mention it.71 Sosylus work was at least known, and possibly available, up to the early
67 H. Hesselbarth, Historisch-kritische Untersuchungen zu dritten Dekade des Titus Livius
(Halle, 1889), 5934; E. Schwartz, Diodorus, RE 1.5 (1903), 663707, at 6889; G. De Sanctis,
Storia dei Romani 3.22 (Florence, 1968), 64750; F. Cssola, Diodoro e la storia romana, ANRW
2.30.1 (1982), 72473, at 763; K.S. Sacks, Diodorus and the First Century (Princeton, 1990),
1312; C.G. Leidl, Appians Annibaike: Aufbau, Darstellungstendenzen, Quellen, ANRW
2.34.1 (1993), 42862, esp. 4579; D. Gaillard (ed. and French trans.), Appien, Livre VII. Le livre
dAnnibal ([Bud] Paris, 1998), xviiixxvii.
68 P. Goukowsky (ed. and French trans.), Appien, Livre VIII, Le livre africain ([Bud] Paris,
2001), lxxixlxxxvii; summarized in Goukowsky, Fragments II, 148, 168.
69 For testimonia, fragments and secondary literature see M. Chassignet (ed. and French
trans.), Lannalistique romaine I. Les annales des pontifes. Lannalistique ancienne (fragments)
([Bud] Paris, 2003), lxxiiiix, 549; H. Beck and U. Walter (ed. and German trans.), Die frhen
rmischen Historiker I: Von Fabius Pictor bis Cn. Gellius ([Texte zur Forschung, Band 76.
Wissenschaftliche Buchgesellschaft] Darmstadt, 20052), 13747.
70 Livy 21.38.25: L. Cincius Alimentus, qui captum se ab Hannibale scribit ex ipso autem
audisse Hannibale = F 10 Chassignet = F 10 Beck-Walter = F 7 Peter = FGrH 810 T6, F5.
71 Diod. 26. fr. a 7a (ed. Goukowsky) (= Exc. 3 Hschel = 4 Walton = FGrH 176 Sosylus T 2):

(
Hschel)
.
Goukowsky, Fragments II, 168, 176 n.15 stresses that Diodorus reliance on or even familiarity
with either Menodotus or Sosylus cannot be assumed on the basis of this notice.

AN UNIDENTIFIED FRAGMENT OF DIODORUS

109

first century A.D., when Diodorus contemporary Cornelius Nepos also implies
familiarity.72 Of the several Greek historians employed in Hannibals retinue Sosylus
appears to have enjoyed a particulary close acquaintance with the general: he
accompanied Hannibal on his campaigns, taught him Greek and wrote a seven-book
history of his achievements, an eye-witness account of the Hannibalic War that would
have provided a corrective to Roman historiographic traditions.73 Sosylus work was
known to Polybius, who censures it as the gossip of the barbers shop; this criticism
need not exclude Polybius use of Sosylus, though this cannot be demonstrated.74 The
sole surviving specimen of Sosylus writing, preserved on a mutilated papyrus at
Wrzburg, is an account of a naval battle fought between the Carthaginian and
Roman-Massiliote fleets, probably at the mouth of the Ebro in 217 B.C., though the
location and date are disputed. This fragment testifies, on the contrary, to his
descriptive precision, detailed information and objectivity, and makes Polybius
judgement seem at best harsh, if not malicious.75 Despite its brevity and damaged
condition, the technical aspects of this battle account, and in particular Sosylus
and the Massiliotes counter-manoeuvre,
description of the Cathaginians
has impressed modern readers. Although the evidence hardly substantiates the view
that Sosylus served Hannibal as a military adviser, as some have conjectured, his
understanding of contemporary warfare appears beyond reproach.76 I shall not press
72 Nepos, Hann. 13.3. It is not possible to discern which details, if any, Nepos might have taken
from Sosylus, or even to be sure that he had direct knowledge of this work, see J. Geiger,
Cornelius Nepos and Ancient Political Biography (Stuttgart, 1985), 10911.
73 Cf. Nepos, Hann. 13.3: huius (sc. Hannibalis) belli gesta multi memoriae prodiderunt, sed ex
his duo, qui cum eo in castris fuerunt simulque vixerunt, quamdiu fortuna passa est, Silenus et
Sosylus Lacedaemonius. atque hoc Sosylo Hannibal litterarum Graecarum usus est doctore, many
writers have handed down a record of his (Hannibals) military achievements, but among these
men there were two who were actually with him on campaign and lived with him as long as
fortune permitted, Silenus and Sosylus the Lacedaemonian. It was this Sosylus whom Hannibal
employed as a teacher of Greek. For older bibliography on Sosylus see Walbank, Commentary,
1.43031. For Greek historians in Hannibals entourage: V. Krings, Les lettres grecque Carthage, in C. Baurain et al. (edd.), Phoinikeia Grammata. Lire et crire en Mditerrane (Namur,
1991), 64968; J. Seibert (n. 15), 1113; D. Briquel, La propagange dHannibal au dbut de la
deuxime guerre punique: remarques sur les fragments de Silnos de Kalakt, in Actas del IV
Congreso Internacional de Estudios Fenicios y Pnicos 1 (Cadiz, 2000), 1237; D. Briquel, Sur un
fragment de Silnos de Cal Act (le songe dHannibal, FGrHist 175, F 8). A propos dun article
rcent, Ktema 29 (2004), 14557; O. Devillers and V. Krings, Le songe dHannibal. Quelques
reflexions sur la tradition litteraire, Pallas 70 (2006), 33746.
74 Polyb. 3.20.5, on which see now V. Krings, La critique de Sosylos chez Polybe III 20, in
G. Schepens and J. Bollanse (edd.), The Shadow of Polybius: Intertextuality as a Research Tool in
Greek Historiography: Proceedings of the International Colloquium, Leuven, 2122 September
2001 ([Studia Hellenistica 42] Leuven, 2005), 22336, with earlier bibliography.
75 FGrH 176 F 1, reproduced with partial French trans. in V. Krings, Carthage et les Grecs
c. 580480 av. J.-C. Textes et histoire (LeidenBostonCologne, 1998), 21729, and again with
detailed commentary in G. Schepens, Die Westgriechen in antiker und modener Universalgeschichte. Kritische berlegungen zum Sosylos-Papyrus, in R. Kinsky (ed.), Diorthoseis.
Beitrge zur Geschichte des Hellenismus und zum Nachleben Alexanders des Groen ([Beitrge zur
Altertumskunde 183] MunichLeipzig, 2004), 73107. See also C. Ferone, Il frammento di
Sosilo sulla battaglia dellEbro del 217 a.C. (F.Gr.Hist. 176 Fr. 1), in M. Capasso (ed.), Papiri
letterari greci e latini (Galatina, 1992), 12739; G. Zecchini, Ancora sul Papiro Wrzburg e su
Sosilo, in B. Kramer et al. (edd.), Akten des 21. internationalen Papyrologenkongresses, Berlin
1995 ([AfP Beiheft 3] StuttgartLeipzig, 1997), 10617.
76 It has been speculated that Sosylus not only taught Hannibal Greek but was also a tactical
adviser, a belief inspired by the statement of Veg. Epit. 3.pr.7 that in Italy Hannibal employed an
unnamed Lacedaemonium doctorem armorum, see e.g. J.F. Lazenby, The Spartan Army

110

PHILIP RANCE

the evidence further, for fear of building a house of cards, but offer L. Cincius
Alimentus and Sosylus for further consideration as the ultimate source for Suda
438, with a marginal preference for the former.
Ludwig-Maximilians-Universitt Mnchen

PHILIP RANCE
prr@fastnet.co.uk

APPENDIX. CITATIONS OF DIODORUS IN THE SUDA

The Suda contains 44 entries derived from Diodorus via extant Excerpta Constantiniana: Diod. 8.7.6 = 308 (4.308.1415) = EV 1.214.1112; Diod. 8.18.1 et 19 =
1271 (4.448.21449.7) = EV 1.215.20216.8; Diod. 13.35.4 = 1871 (1.168.1315) =
1596
EV 1.232.24; Diod. 29.17.1 (Alder 5 Ind. Auct. 75 errore 19.17.1) =
(2.147.246) = EL 400.267; Diod. 19.81.34 = 429 (2.41.210) = EV 1.249.1221;
Diod. 21.17.1 = 600 (4.553.5) = EV 1.254.19; Diod. 21.17.13 = 602 (4.554.228) =
EV 1.254.19255.17; Diod. 21.17.4 [errore 21.7.14 in Adler] = 214 (3.17.2031) =
EV 1.275.1830; Diod. 23.16 = 3393 (2.442.1416) et 442 (4.41.256) = EL
406.710; Diod. 24.3.2 = 341 (4.33.1114) = EV 1.260.1016; Diod. 28.15 [errore
Adler 38.15] = 2657 (2.385.212) = EL 398.89; Diod. 29.4 = 3602 (1.325.267) =
EL 398.278; Diod. 29.9 = 2642 (1.234.14) = EL 399.33400.2; Diod. 29.17.1 [omit
Adler] = 1596 (2.147) = EL 399401; Diod. 29.11 = 466 (4.668.78) = EL 400.78;
Diod. 29.12 = 676 (3.52.24) = EL 400.1820; Diod. 29.33 = 457 (4.819.1113) =
EL 401.1618; Diod. 30.2 = 1763 (1.157.1114) = EL 401.31402.1; Diod. 31.5.3 =
522 (4.49.57) = EL 402.235; Diod. 31.7.2 = 3088 (1.277.13) = EL 403.57;
Diod. 32.6 = 444 (4.42.17) = EL 404.312; Diod. 32.6.1 = 1935 (1.173.279) et
798 (2.75.1315) = EL 404.2731; Diod. 32.6.4 = 2246 (2.350.223) = EL 405.68;
Diod. 33.5 = 2440 (1.217.235) et 4611 (1.431.45) = EL 405.203, 268; Diod.
33.16 = 217 (2.203.245) = EL 406.1011; Diod. 33.28ab = 316 (1.33.1820) et
1835 (1.165.56) et 121 (1.455.1314) et 8 (1.503.12) et 2673 (2.386.1812) et
3533 (2.453.2930) et
3786 (2.477.1416) et
111 (2.609.1517) et
865
(3.302.1922) = EL 406.1721, 32407.3, 610, 1419; Diod. 34/5.38.1 = 462
2362 (4.201.1617) = EV
(2.47.312) = EV 1.312.1718; Diod. 34/5.34 =
1.313.1317; Diod. 36.15.23 = 4274 (1.394.1012) et 2759 (2.393.89) et 787
1495 (4.467.23) = EL 408.202, 2932; Diod. 40.1.3 =
2319
(3.59.78) et
(1.208.68) et 405 (4.531.212) = EL 409.235, 313.
Another 11 entries not found in the surviving volumes of Excerpta can be
identified from the extant books of Diodorus: Diod. 13.67.2 (Adler 5 Ind. Auct. 75
errore 62.2) = 3559 (1.321.256); Diod. 14.10.3 = a 2062 (1.186.12) (falso Polyb. fr.
106); Diod. 14.20.3 = 4040 (1.369.89); Diod. 14.23.1 = 354 (4.659.27660.2);
Diod. 15.53.4 = 369 (2.36.35); Diod. 15.84.2 = 2067 (2.339.89); Diod. 17.82.7 =
2617 (1.232.45); Diod. 17.85.5 = 2818 (4.234.45); Diod. 19.96.4 = 3722
(1.335.89); Diod. 20.11.5 = 444 (4.42.1415); Diod. 26.2 = 240 (4.801.46).

(Warminster, 1985), 170; G. Zecchini (n. 75), 10634; Goukowsky, Fragments II, 105, n. 144, 141,
n. 38, 176, n. 16. E.L. Wheeler, The Hoplomachoi and the legend of Spartan drillmasters,
Chiron 13 (1983), 120, at 12, 1516, casts doubt on this thesis, arguing that although Vegetius
probably does allude to Sosylus, the Spartans military role is a mistaken late deduction by
Vegetius himself or his source.

AN UNIDENTIFIED FRAGMENT OF DIODORUS

111

A further 9 entries name Diodorus, but come from sections of his work preserved
in neither the Excerpta nor the direct tradition: Diod. fr. 5 = 2668 (1.236.234); fr. 6
= 1471 (2.295.1617); fr. 7 = 1521 (2.299.1719); fr. 8 = 1803 (3.135.234); fr. 9 =
1021 (4.427.23); fr. 10 = 1544 (4.470.1314); fr. 11 = 586 (4.676.246); fr. 12 =
9 (4.780.57); fr. 13 = 538 (4.828.1718).
There are 12 suspected fragments: Suda 741 (1.71.67) (vel Polyb.); Suda 1961
(1.176.1112); Suda 2612 (1.231.224); Suda 2925 (1.264.1819); Suda 1368
(2.129.213) (vel Polyb.); Suda 19 (2.189.14) (vel Polyb.); Suda 3524 (2.453.78);
Suda 179 (2.532.8) (vel Cass. D.); Suda 647 (3.50.25) (falso Polyb. fr. 169; Diod.
vel Cass. D.); Suda 65 (3.440.13) (vel Cass. D.); Suda 2231 (4.191.10); Suda 556
(4.747.34).

Classical Quarterly 59.1 112124 (2009) Printed in Great Britain

doi:10.1017/S00098388090000081

112
STOIC AND POSIDONIAN THOUGHT ON THE IMMORTALITY OF
SOUL
A.E.
JU

STOIC AND POSIDONIAN THOUGHT ON THE

IMMORTALITY OF SOUL
What did Posidonius mean by the immortality of soul? Was his reference to the world
soul only or also to individual souls, including human souls? Is Posidonius conception of immortal souls genuinely Stoic? In attempting to answer these questions, this
article initially sets out to elucidate Posidonius response to Platos assertion of the
souls immortality advanced at Phaedrus 245c. My discussion further relates a Stoic
and possibly a Posidonian conception of souls perishability to a line of Stoic thought
Posidonius would have inherited regarding destruction, in which Chrysippus notably
developed an account of the perishability of all qualified souls. In outlining a
plausible theory by which Posidonius might have held both to souls immortality and
their perishing, this work attempts to reconstruct the meaning Posidonius might have
attached to these terms in their likely context of intellectual exchange with Stoicisms
critics, notably Cicero. Attention to Posidonius arguments in this quasi-polemical
context suggests the Stoic orthodoxy of his views, further demonstrating how
Stoicism was capable of defusing a number of apparent contradictions between the
postulates of souls immortality and perishing through positing sophisticated distinctions between these terms. This paper further advances the claim that Posidonius
interpretation of Platos argument makes sense within a context of his reading of
Platos dialogue. It is likely that Posidonius was motivated to draw on Plato in the
Phaedrus in order to equate Platos all soul with whatever is self-moving, yielding a
Posidonian characterization of deathlessness as the whole of self-moving soul.
I. IMMORTAL SOULS
Hermias, in a passage of his commentary on Platos Phaedrus, reports Posidonius
to introduce
interpretation of Phaedrus 245c, where Plato uses the words
his argument for the immortality of soul. In offering two contrasting interpretations
of Platos argument, Hermias singles out Posidonius as representative of the group
who took the words to refer to the world soul only, in contrast with Harpocration,
who supposedly took the term to refer to absolutely all soul. Festugires view aside,1
Stoic scholarship has broadly advanced two views of this crux. Hoven suggests that
Posidonius reading represents a disinterested attempt to elucidate Plato without
guidance from his own philosophy. Edelstein infers, contrarily, that Posidonius
adopted his interpretation on the strength of his own belief in the immortality of the
world soul only. Kidd largely concurs with this latter view, without finding adequate
basis in Hermias text for the inference.2 In a competing attestation, Div. 1.64,
1 A.-J. Festugire, Platon et lOrient, RevPhil 3.21 (1947), 545, at 21 judged all soul to refer
to soul collectively, and, in siding with Posidonius, took the ensuing argument of Plato to refer
exclusively to the world soul; others such as T.M. Robinson, Platos Psychology (Toronto, 19952),
11118 reject this interpretation.
2 R. Hoven, Stocisme et stociens face au problme de lau-del (Paris, 1971), 62; L. Edelstein,
The philosophical system of Posidonius, AJPh 57 (1936), 286325, at 300, n. 58; I.G. Kidd (ed.),
Posidonius: the Translation of the Fragments (Cambridge, 1999), 143, Posidonius: the Commentary (2 vols; Cambridge, 1988), 2.97981.

STOIC AND P OSIDONIAN THOUGHT ON THE IMMORTALITY OF SOUL

113

however, Cicero reports that, in discussing the possibility of human dreams representing means of divination, Posidonius spoke of immortal souls. Unless we distrust
Cicero, it becomes difficult to infer both that Posidonius held to a belief that the
world soul alone is immortal and that his interpretation of Plato was influenced by
such a belief. In the light of Ciceros reference, the question of Posidonius own view
on the topic needs to be scrutinized further. Therefore I address this question first,
before discussing Posidonius interpretation of Platos argument.
Cicero, Div. 1.64 reads:
Sed tribus modis censet deorum adpulsu homines somniare, uno quod provideat animus ipse per
sese, quippe qui deorum cognatione teneatur, altero quod plenus aer sit inmortalium animorum,
in quibus tamquam insignitae notae veritatis appareant, tertio quod ipsi di cum dormientibus
conloquantur.
He proposes three ways in which men dream through divine impact: the mind of its own nature
foresees, inasmuch as it is imbued with kinship with the gods; the air is full of immortal souls, in
which appear, as it were, clear marks of truth; the gods themselves speak with men who are
asleep.
(Cic. Div. 1.64; fr. 108 EK, part).3

This Ciceronian attestation has given rise to extensive debate among scholars seeking
to clarify Posidonius apparent claim: the air is full of immortal souls. Edelstein
takes the expression immortal souls to refer to the fixed stars, arguing further that it
is impossible to conclude from this assertion that Posidonius believed the human
soul to be immortal.4 But in my view such a conclusion is not entirely impossible. In
Ciceros testimony above, Posidonius description of souls as immortal is not further
specified, nor, to my knowledge, is there any evidence that Posidonius considered
human souls mortal. Further, the words the air seem to rule out the identification of
souls as the fixed stars; for the Stoics normally understood the sphere of the fixed
stars [to be] created not in the airy but in the fiery or ethereal region. Posidonius
likewise associated the stars with the heaven, which he elsewhere called the outermost circumference, and which he and Chrysippus nominated as the dwelling-place
of the cosmic deity and celestial gods.5
In seeking to identify the souls that are said by Posidonius to be in the air, we
may need briefly to advert to the doxographical section of Sextus Empiricus, Math.
9.734.

They [souls], having quitted the sphere of the sun, dwell in the region below the moon, and there
because of the pureness of the air they remain for a long time, and for their nutrition they use the
vapour rising from the earth, as do the rest of the stars also, and in those regions they have
nothing to dissolve them. If, therefore, the souls remain, they come to be the same as daemons;
and if there are daemons, then we must say that gods too exist, their existence being by no means
hindered by the preconception about the legendary doings in Hades. (Sext. Emp. Math. 9.734)

3 For scholars discussion of this testimony, see Edelstein (n. 2), 300; Hoven (n. 2), 58; M.
Dragona-Monachou, The Stoic Arguments for the Existence and the Providence of the Gods
(Athens, 1976), 1734; Kidd (n. 2), 1.42832.
4 Edelstein (n. 2), 300, n. 58.
5 Diog. Laert. 7.1378, 7.144; Simpl. In Ar. Cael. 4.3.310b1 (fr. 93a EK).

114

A.E. JU

In the previous section 9.712, the same source, against Epicurus, insists upon the
existence of souls in the extra-terrestrial region, in asserting that, after separating
from their bodies, they neither move downwards nor are dispersed but rather soar
lightly into the upper region. The grounds for this argumentation rest both on the
souls causal effect upon themselves through their capacity of holding themselves
together, and on their pneumatic nature, composed as they are out of airy and fiery
fine particles.6 In the present section just quoted, the passages author further adduces
arguments in favour of the souls survival, citing in particular the airs pureness in the
region below the moon.
Two points should be taken into account in considering the nature of these souls.
First, the authors demonstrations deploy a wide range of Stoic terms and ideas, for
instance describing the souls as deriving nutrition from the vapour rising from the
earth, as do the rest of the stars,7 and even playing on the etymology of the word
daemon in the sentence if, therefore, the souls remain (
), they come to be
).8 Further, the souls considered in the Sextus
the same as daemons (
passage are likely to refer to human souls surviving death, or more specifically to
heroes; for only this assumption justifies the contention that the souls residing in the
upper region for a long time come to be the same as daemons.
This material preserved by Sextus, whoever his Stoic source, provides persuasive
grounds for believing that the expression immortal souls in the air, as suggested at
Cicero, Div. 1.64, refers to daemons, that is, to divine souls intermediate between the
stars and human beings, or to heroes, signifying the souls of dead virtuous men who
have survived death. Perhaps both references are meant, seeing that for the author of
the Sextus passage the expression the region below the moon as the location of
surviving human souls and daemons best parallels the air in Ciceros report. The
author is clear that the souls reside in this region because of the pureness of the air,
which presumably facilitates their becoming the same as daemons. Kidd, in a passage
of his commentary, makes a similar point to mine regarding Ciceros report; but this
makes it hard to understand how and why Kidd elsewhere insists that Posidonius did
believe in the sole immortality of the world soul.9 Posidonius assertion in Ciceros
report sufficiently confirms that he admitted the immortality of individual souls, in
addition to that of the world soul.
There is a question, though, whether this Posidonian conception of the souls
immortality can legitimately be taken as Stoic, since a range of evidence presented
below rather attests a Stoic conception of the perishability of souls. The Stoics,
according to Diogenes Laertius 7.156, infer, first, that soul is a body,

and, then, that it survives death; but it is perishable, though the soul of the universe, of which
those [souls] in animals are parts, is imperishable.
(Diog. Laert. 7.156)

Sext. Emp. Math. 9.712; see also Ach. Tat. Intr. in Arat.13 (fr.149 EK).
The discussion of the stars nutrition (apart from that of the sun and moon) at 9.73 stands
parallel to Diog. Laert. 7.145 (frr. 9, 10, 17 EK) and Cic. Nat. D. 2.118.
8 The word
, like
, is an original Stoic term. Sext. Emp. Math. 9.74 can
serve as evidence for a Stoic, or a Posidonian, etymology of daemon, in addition to those at
Macr. Sat. 1.23.7 (fr. 24 EK). For the Stoic concepts of daemons and heroes, see Diog. Laert.
7.151.
9 Kidd (n. 2), 1.4301, 2.981.
7

STOIC AND P OSIDONIAN THOUGHT ON THE IMMORTALITY OF SOUL

115

That is, for the Stoics, the world soul is identical to the whole of soul, whereas
), as they were
individual souls, such as those of animals, constitute its parts (
unified with the world soul during the conflagration, and are now, as Diogenes
) of this soul, a formuelsewhere reports, a fragment or an offshoot (
lation to which Posidonius seems to have taken no significant exception.10
Further, Diogenes doxography ascribes the idea of the soul surviving death but
later perishing to the Stoics in general. In doing so, it draws our attention to two
points. First, perishable is a standard term used by Stoic writers to describe the
nature of all individual souls, in implicit contrast to the imperishable world soul.
Second, the soul Diogenes here speaks of refers to individual souls generally, but
more specifically to human souls, whether rational or irrational; for, to the best of my
knowledge, Stoics as early as Zeno did not describe the souls of non-rational animals,
the stars or daemons as surviving death.
Diogenes doxography begins with general statements about souls advanced by
the Stoics collectively, but afterwards cites Cleanthes as maintaining that all souls
survive up to the conflagration. Chrysippus, meanwhile, Diogenes asserts, holds
this to be true only of the souls of sages.11 Eusebius too, in his doxography of Stoic
psychology, reports the same Chrysippean idea as Diogenes does, according to which,
among the souls separated from their bodies, not all but only human souls survive on
their own for a certain time; those of non-rational animals perish at once. Chrysippus
appears to have maintained further that not even all human souls but only those of
virtuous men survive up to the conflagration; fools souls do so only for a limited
period, perishing sooner.12
The above evidence has two major implications. First, the Stoic assertion of souls
perishability should not be taken to mean that the souls are destroyed in an unqualified sense, but that they undergo a sort of natural change which the Stoics called
broadly resolution, by which they are unified with the imperishable world soul so as
to become its parts. Second, while believing individual souls to experience this resolution generally, Chrysippus, unlike Cleanthes, discriminated between the souls of
sages and fools with regard to the mode of destruction. That is, for Chrysippus, sages
souls undergo the conflagration, whereas those of fools are not subject to this particular form of resolution, rather experiencing what might be called mere resolution.
To make these points clearer, we should remember here that in Stoic physics
resolution holds a meaning far broader than that of conflagration, referring in
Chrysippus physics to a compounds dissolution into its components (for instance, an
eggs dissolution into yolk, white and shell, etc). The Stoics further explained the
dissolution of the world and its individuals into the four elements (fire, air, water and
earth) as an instance of this kind of destruction, indicating more specifically
rarefaction or the reverse of condensation in the elements reciprocal changes:
10

Diog. Laert. 7.143 (fr. 99a EK). For the Stoic conception of the conflagration, see n. 14.
Diog. Laert. 7.157. It seems uncertain whether Cleanthes expression all in Diogenes
doxography refers to all animals souls or to all human souls; perhaps the latter is meant.
12 Euseb. Pr. ev. 15.20.6 (SVF 2.809): They say that the soul is subject to generation and
destruction. When separated from the body, however, it does not perish at once, but survives on
its own for certain times, the soul of the virtuous up to the dissolution of everything into fire, that
of fools only for a certain definite time. By the survival of souls they mean that we ourselves
survive as souls separated from bodies and changed into the lesser substance of the soul, while
the souls of non-rational animals perish along with their bodies (trans. by LS 53W). Cf. A.A.
Long and D.N. Sedley (edd.), The Hellenistic Philosophers (2 vols; Cambridge, 1987), 2.3201;
Hoven (n. 2), 62; J.M. Rist, Stoic Philosophy (Cambridge, 1969), 25661.
11

116

A.E. JU

(Reciprocally, the dissolution and

diffusion of earth, the first diffusion into water, and the second from water into air,
and the third and last into fire).13 This last phase is still not identical with the conflagration, since ordinarily the rarefaction of air into fire occurs within a cyclical
sequence proceeding immediately to condensation in its next stage, in which the world
is partly rearranged.
The conflagration, on the other hand, is understood by the Stoics as the devastation of the worlds arrangement as such, so that all, namely the whole of the world,
is resolved into, and unified with, a single element fire.14 This process also represents
a sort of rarefaction, but, unlike the normal referent of the term rarefaction, the
conflagration is unidirectional and does not at once accompany condensation, rather
determining the worlds periodic destruction into fire at very long intervals of
everlasting recurrence. In Stoic physics the conflagration is in this light a particular
form of resolution, applying above all to the destruction of the elements, the world,
the gods other than Zeus, and the souls of sages.
The question arises why Chrysippus believed that the souls of sages survived at all
up to the conflagration, and what sense this made for him as a psychological or ethical
doctrine, as well as a physical one. This question is controversial and I attempt no
direct treatment of it here. But the extant Stoic evidence, however meagre, allows us to
conjecture that, in Stoicism, among the souls separated from their bodies and
changed into their lesser substance those of sages alone are able to survive up to the
conflagration on the strength of their superiority both in intelligence and in pneumatic constitution. Presumably part of the answer to the above question lies in the
Stoics treatment of Platos Phaedo as a canonical Socratic text, an attitude which
meant that they had to extract from it lessons which could be made compatible with
Stoicism. That is, total indestructibility of individual souls would have clashed with
the conflagration theory, but limited indestructibility as a reward for virtue could be
made fully consistent with Stoicism.15
This Chrysippean idea of the human souls destruction, closely linked with his
physics of the conflagration, probably represented a standard Stoic doctrine. Little
evidence remains, however, as to how the post-Chrysippean Stoics responded to
Chrysippus on the matter of souls perishability. Inasmuch as Panaetius, siding with
Boethus of Sidon and Diogenes of Babylon, denied the theory of the conflagration, it
seems reasonable to suppose that Panaetius had no motive for following Chrysippus
in every detail, preferring to use his own syllogisms in seeking to prove souls
perishability.16 Posidonius at least differed from these three Stoics in abiding by the
13

Stob. Ecl. 1.129.2130.13 (SVF 2.413, part); see also Diog. Laert. 7.1367, 7.142.
Diog. Laert. 7.134 (fr. 5 EK); Sen. Ep. 9.16; Euseb. Pr. ev. 15.18.2 (SVF 2.596); Alexander
Lycopolis 19.24 (LS 46I); Plut. Comm. not. 1075d (SVF 1.510); Cleom. De motu 1.1 (SVF
2.537); see also SVF 2.61332.
15
I am grateful for, and accept, the suggestion of D. Sedley, tending to confirm the attribution
above. The pneuma constituting the souls of sages is fierier, finer and more in a hierarchy of
tension than that of fools souls; and after separating from their bodies they reside in the upper
region as an intelligent and fiery pneuma through their capacity of holding themselves together,
sharing also in some instances in the ethereal substance which Zeus entirely occupies at the
conflagration. Cf. n. 6; Stob. Ecl. 1.213.1521 (SVF 1.120); Plut. Comm. not. 1077e; Sext. Emp.
Math. 9.86; Diog. Laert. 7.1389.
16 Cic. Nat. D. 2.118; Philo, Aet. mund. 2.497 M (4.96.19 Cohn) (LS 46P; fr. 99b EK, part);
M. van Straaten, Panaetii Rhodii fragmenta (Leiden, 19623), 649; Long and Sedley (n. 12), 2.277.
For Panaetius syllogisms about the souls perishability, see Cic. Tusc. 1.79.
14

STOIC AND P OSIDONIAN THOUGHT ON THE IMMORTALITY OF SOUL

117

earliest Stoic doctrine of the worlds periodic destruction at the conflagration.17

Further, as Arius Didymus testimony shows, Posidonius, concurring also with
mainstream Stoics,18 held to a version of the theory classifying destruction into four
types19: division, alteration, fusion and resolution, in Posidonius wording an
out-and-out-change,20 attributing alteration to substance only and the other three
modes to qualified individuals. Arius doxography, however, provides no information
as to Posidonius examples. It thus remains a matter of speculation what forms of
destruction he understood as applying to individual souls, and particularly to human
souls.
This lack of direct evidence prevents us from fully recovering the terms of the
debate inherited by Posidonius from Chrysippus on the topic of souls perishability,
and even whether such a debate existed at all. But in so far as Posidonius adhered to
the standard Stoic doctrine of destruction as noted above, it is possible to suppose
that he related individual souls to resolution, holding to an idea of perishable souls.
On this supposition, however, in addition to the question of Posidonius Stoic
orthodoxy in insisting on souls immortality, the compatibility for him between the
postulates of souls immortality and perishing also becomes a matter for discussion,
especially when we bear in mind Ciceros objection to a Stoic belief in perishable souls
on the basis of the immortality of soul.
Cicero, at Tusc. 1.18 and 1.778, while siding with Plato, indeed makes a case for
the immortality of soul, categorizing those who define death as a separation of soul
17 Scholarship is now practically unanimous that Posidonius concurred with the early Stoics
and not with his master Panaetius regarding the conflagration. Cf. Diog. Laert. 7.142 (fr. 13 EK);
At. Plac. 2.9.3 (Stob. Ecl. 1.18.4b; DG p. 338.17) (fr. 97ab EK).
18 Diog. Laert. 7.141. That the world and its parts are subject to destruction is standard Stoic
doctrine. Diogenes Laertius doxography shows that the Stoics adduced a syllogism proving the
susceptibility of the world to this change.
19 Ar. Did. Epitome fr. 27 (Stob. Ecl. 1.20.7, 1.177.20 W; DG p. 462) (fr. 96 EK, part):
Posidonius says that there are four kinds of destruction and generation from being to being. For,
they recognized that there was no such thing as generation from, or destruction into, non-being,
as we said before. But of change into being he says that one kind is by division (
),
one by alteration (
), one by fusion (
), and one an out-and-out
change (
), which they call by resolution (
); of these, that by alteration belongs to the substance, while the other three belong to the so-called qualified individuals,
which come to occupy the substance. For the discussion of the testimony, see Long and Sedley
(n. 12), 1.1723; Kidd (n. 2), 1.38490. Arius doxography represents an important source for
Posidonius conception of generation and destruction. Posidonius classification concerns less
change as such than those changes by which a things identity can be lost, as Long and Sedley (see
above) point out. Supposing that Arius quotation is from a Posidonian source, possibly the word
they referred for Posidonius to mainstream Stoics, and the word we suggests that he concurred
with them on the definition of change. I will not discuss the question why Posidonius gave such a
quadripartition, different from the Chrysippean (
) as cited
at Stob. Ecl. 1.17.4, p. 154 W (SVF 2.471), and also from the Stoic tripartition (
) as cited at Philo, Aet. mund. 79.
20 As Arius testimony (n. 19) shows, Posidonius called one sort of destruction an out-and-out
change (
). It is uncertain why he introduced this new expression, on which Arius
does not specifically comment in his doxography. Bearing in mind both the Stoic conception of
conflagration (
) (see n. 14) and the ensuing clause of the testimony which they call
by resolution (
) , Posidonius might not have meant by that expression the conflagration exclusively, rather implying the resolution of qualified individuals more broadly. Bernard
Collette, however, has suggested (in a Cambridge seminar, 2006) that Posidonius expression an
out-and-out change in Arius doxography can be regarded as direct evidence for Posidonius
admitting the conflagration, a point Kidd (n. 2), 1.387 fails to pick up. However, Kidds caution
may be well placed, on my reading of Posidonius expression as alluded to above.

118

A.E. JU

from body into three: some hold that it is at once dispersed in space, others that it
survives a long time, and others that it survives eternally. Within this tripartite
grouping Cicero places the Stoics between the Epicureans and Platonists, marking off
all three from Dicaearchus who, in Ciceros view, argued most incisively against the
souls immortality. Cicero remarks of the Stoics that they, on the other hand, give us a
generous lease of life, as though to make us crows, testifying to a certain scepticism as
to whether the soul can indeed survive a long time but not eternally. Conceivably, this
Stoic view on perishable souls represented a major stumbling block to Ciceros
acceptance of Stoicism, since he himself speaks of it as the point of greatest difficulty
in the whole problem.
There seems little reason to deny that mainstream Stoics, or even those contemporary with Cicero, would have been forced to respond to criticism of this kind. Little
evidence of the Stoic responses remains, however; and Cicero himself informs us no
farther than reporting Panaetius syllogisms about souls perishability.21 But from
such evidence as exists we can highlight at least a few aspects of these debates. Consideration of Stoics debates presented below suggests that a conception of immortal
souls is as Stoic as a conception of perishable souls, and further determines Posidonius fundamental adherence to Stoic orthodoxy on this topic.
In considering the grounds on which mainstream Stoics and Posidonius might have
argued for the souls immortality and perishing, we should particularly remember that
no Greek thinkers outside the Stoic school marked off immortal and imperishable
from each other with regard either to soul or to god. The Phaedos final argument, for
instance, appears to take immortality to entail imperishability, though these two
words are not semantically identical, hereby proving the soul to be both immortal and
imperishable.22 The Stoics are, however, well known for employing imperishable
exclusively for their supreme cosmic deity Zeus, at the same time using perishable of
the rest of gods and souls as they did of the world.23 They employed immortal,
however, for all of these without discrimination, as a range of evidence shows.24 We
can therefore legitimately suppose that the founding fathers of the Stoic school made
certain conceptual distinctions between immortal and imperishable, inferring, for
instance, that the souls are immortal but perishable. This Stoic terminology is exactly
what Cicero, and later also Plutarch,25 seized on to launch one of their strongest
21 Cic. Tusc. 1.7980. It is uncertain why Cicero argues especially against Panaetius, but
perhaps he wanted to show how Panaetius, who revered Plato most among philosophers,
neglected the fact that, when Plato spoke of the eternity of souls, this pertained not to the souls
irrational parts, but to the mind which is always distant from disorderly impulse.
22 Pl. Phd. 105e10107a1; see also Phdr. 245c5246a2. Plato, at Phd. 106d46, applies this
combined vocabulary also to god and the form of life itself. Platos precise position in the
Phaedos final argument on immortality and imperishability is more complex than I indicate,
though this does not affect points I have made above. Cf. D. Gallop (ed.), Plato: Phaedo (Oxford,
19885), 21622; D. Frede, The final proof of the Immortality of the Soul in Platos Phaedo
102a107a, Phronesis 23 (1978), 2741; R. Woolf, The practice of a philosopher, OSAP 26
(2004).
23 For the Stoic use of
for Zeus, see Diog. Laert. 7.137, 7.134; Plut. St. rep. 1051ef,
Comm. not. 1077e; of
for the world and the soul, see nn. 12, 18.
24 For the Stoic use of
for god(s), see n. 25; Diog. Laert. 7.147; Cic. Nat. D. 1.123 (fr.
22a EK), 2.45; Cleanthes, Hymn. (Stob. 1.25.327.4); Sext. Emp. Math. 9.85; Calc. In Tim. 293;
for the world, see Plut. St. rep. 1052cd.
25 Plut. Comm. not. 1075c. This Plutarchean attestation, for all its polemicism, goes to show
that mainstream Stoics asserted that god is not mortal but perishable (
). Plutarch here took mortal to mean nothing other than perishable and thus
argued that the Stoics were inescapably caught in a dilemma, violating the common conception
of god.

STOIC AND P OSIDONIAN THOUGHT ON THE IMMORTALITY OF SOUL

119

objections to Stoicism. But it is easy to see this inference as relying both on an

originally Platonic definition of death26 and on a Stoic conception of destruction, in a
few regards expanded on below.
The Stoics most probably understood death as a form of division, namely
separation of soul and body,27 through which any ensouled being loses its own
identity.28 To the extent that they adhered to this definition of the term, they could say
that a man dies but a soul or a god does not. That is, according to the Stoics, it is we
and not our souls who are subject to death; hence we ourselves survive as souls, as
Chrysippus says, separated from bodies and changed into the lesser substance of the
soul, as cited by Eusebius in the passage above.29 Further, for the Stoics, because the
gods are divine, and also because they experience no separation of soul from body,
despite being composed of these parts, they are correctly called immortal or
deathless.30 Chrysippus, in Book 1 of On Providence, applied this exact attribute to
the world, in asserting that the world soul is not separated, but grows continuously
until it has completely used up its matter on itself ; hence the world must not be said
to die (
) (Plut. St. rep. 1052cd).
The question arises how far the very focus on immortality in this Stoic material
betrays a preoccupation with Platonic concerns as to the souls immortality advanced
in the passages of the Phaedo. It is uncertain whether mainstream Stoics or Posidonius worked with the relevant passages of the dialogue. But the Stoics conception of
the souls immortality is in no sense a deviation from the language of Platonism, in
that they adhered to a Platonic definition of death. Supposing the allegedly
Pythagorean origin of the Platonic idea expounded in the passages, it would even
seem possible that the Stoics rather wanted to adhere to a core Pythagorean idea of
the souls immortality.31
Nevertheless, the Stoics conception of immortal and perishable souls seems to
indicate where they diverged from the Phaedos final argument. It remains conceivable
that the validity of this argument depends on the condition that the deathless
should also be indestructible.32 The Stoics, however, may have found this condition
implausible, perhaps on the basis of their discrimination between death and the other
forms of destruction as previously considered.33 Since, as noted, for the Stoics death
refers narrowly to a form of division (into soul and body) befalling any ensouled
being, a soulless thing, for instance an egg, is not said to die in Stoic terms, but is
taken by them to be destroyed, whether by mere division (into yolk and white), by
fusion (for instance, into a cake), or by resolution (into its components). Further, as
indicated, a soul or a god is not said to die either, but is taken to be destroyed by

26

Cf. Pl. Phd. 64c29, 67a, 105d13e9; Cic. Tusc. 1.18.

Cf. Sext. Emp. Math. 7.234; Plut. St. rep. 1052cd; Nem. De nat. hom. 81.610; Cic. Tusc.
1.18; see also Long and Sedley (n. 12), 1.173.
28 In Arius Didymus testimony (n. 19), of the three types attributed by Posidonius to qualified
individuals, division describes a change befalling any unified body, in so far as this body is
something qualified; that is, when this change occurs in a compound, the body loses its identity.
29 Cf. n. 12.
30 Cf. Long and Sedley (n. 12), 2.454. Frede (n. 22), 30 states that the word
is as
ambiguous as the English word immortal , since it designates not only deathlessness but also
everlastingness. I agree; the Stoic term immortal is perhaps best rendered as deathless.
31 I am grateful for, and accept, the suggestions of M. Schofield in a Cambridge seminar
tending to confirm this attribution.
32 Frede (n. 22), 30.
33 Cf. n. 19.
27

120

A.E. JU

resolution, or more specifically by conflagration. The Stoics point is clear, namely

that in Stoic classificatory terms deathless does not necessarily entail indestructible;
and that, in so far as a thing is something qualified, though not experiencing death,
this thing can undergo other forms of destruction.
In this characterization of the Stoic theory of destruction, Stoicism shows itself
capable of defusing all manner of apparent contradictions between the postulates of
the souls immortality and perishing by positing sophisticated distinctions between
these terms. There seems little basis for doubting that Posidonius conception of souls
immortality was genuinely Stoic; and that, in claiming that the air is full of immortal
souls, he was not deviating from Chrysippus or even from Plato. Bearing in mind
Ciceros knowledge of Stoicism in general, we must assume that Cicero in this respect
was aware of the Stoic terminology, and rejected it.
II. SELF-MOVING SOUL
We can now profitably turn our attention to Posidonius interpretation of Phaedrus
245c, as reported in Hermias testimony (though my examination here will necessarily
be glancing). Platos own lemma reads:

[]

All soul is immortal; for what is self-moving is immortal; but what is itself moved by something
else, imparting motion, ceases its motion, and therefore ceases to live. Only what moves itself
never ceases its motion, inasmuch as it cannot abandon itself; moreover, this self-mover is the
source and the first principle of motion for all other things that are moved. Now a first principle
cannot come into being. [] It is as impossible that it should be destroyed as that it should come
into being; were it otherwise, the whole heaven and the whole of becoming would collapse, stop,
and never again have any other source of motion to bring it back into being.
(Pl. Phdr. 245c5e3)34

Hermias states:

34 For scholars debates over this passage, see P. Frutiger, Les Mythes de Platon (Paris, 1930),
1304; L. Robin, Platon. Phdre (Paris, 1933); J.B. Skemp, The Theory of Motion in Platos Later
Dialogues (Cambridge, 1942), 3, n. 1; Festugire (n. 1), 21; C. Diano, Quod semper movetur
aeternum est, PP 2 (1947), 18992; R. Hackforth, Platos Phaedrus (Cambridge, 1972), 64, n. 3;
Robinson (n. 1), 11118; M. Trapp, Platos Phaedrus in second century Greek literature, in D.A.
Russell (ed.), Antonine Literature (Oxford, 1990), 14173; R. Bett, Immortality and the nature of
the soul in the Phaedrus, in G. Fine (ed.), Plato (Oxford, 2000), 90731. The sentence at 245c5 is
problematic. I follow the reading of the papyrus (P.Oxy. 1017)
(accepted by Robin,
Bignone, Mller, Ross and Ackrill), and not the reading of OCT
(accepted by
Hackforth, Diano and Robinson). I follow Hermias reading of
(accepted by Hackforth).
The most serviceable translation of Platos words
is perhaps all soul, but this
wording does not answer the question how to take the expression.

STOIC AND P OSIDONIAN THOUGHT ON THE IMMORTALITY OF SOUL

121

We must first ask what kind of soul he [Plato] means. Some thought that his argument referred
to the world soul only, because he said all and a little later added that were it otherwise, the
whole heaven and the whole of becoming would collapse and stop; the Stoic Posidonius is one
of these. Others thought that it referred to absolutely all [soul], including the souls of ants and
flies; Harpocration is one of these; he understands all as applying to all soul.
(Herm. In Phdr. ad 245c; fr. 290 EK)35

A central difficulty in probing Posidonius equation in Hermias report of all soul

with the world soul lies in the entire lack of context. Even so, from what we know of
Posidonius it is not likely that he adopted this equation on the basis of his belief in
the immortality of the world soul only. Nor did his reflection centre on the slightly
different concept of the world soul as advanced by Plato in the Timaeus; for, as
Posidonius presumably knew, the world soul in the Timaeus represents one particular
kind of rational soul, alongside other divine souls and the rational part of human
souls.36 Instead, there is reason to believe that Posidonius took account of both the
myth and the wider context of the Phaedrus, first because Hermias quotes a relevant
section of Platos argument in introducing Posidonius interpretation, and second
because Posidonius elsewhere himself made use of Platos figure in the myth of two
horses drawn by a charioteer in describing the souls two irrational parts.37
Assuming Posidonius familiarity with the dialogues context, it remains conceivable that he may not have neglected the fact that Platos argument for the immortality
of soul is meant to be applicable to individual souls, however indirectly.38 Further,
while Hermias insisted that Platos argument is limited to the soul of god and the
rational part of the human soul,39 Posidonius at least does not seem to have attended
to this sort of limitation. Possibly he thought that Platos focus lies elsewhere; the
emphasis in Posidonius eyes seems to fall much more heavily on the significance of
Platos introducing his argument with the word all, as Hermias report shows. It
seems possible, then, that Posidonius inferred all soul to refer for Plato to the whole
of soul collectively, including all individual souls. In addition, the principal question
in Stoic scholarship on Posidonius psychology is whether, and if so in what sense, he
himself acknowledged there to be parts to the soul; whatever his actual view on the
topic, though, this does not seem to have affected his interpretation of all soul.
But, even if this conjecture is correct, the question why Posidonius in his interpretation employed the expression the world soul needs to be answered. Instructively, in
Stoic physics, as already noted, the world soul is identical to the whole of soul
subsuming all individual souls, whereas individual souls are its parts or offshoots.40 In
this characterization of the Stoic view on soul, it would seem possible that in
interpreting all soul Posidonius recruited Plato to the Stoic side; these words become,
35 For scholars debate about Hermias text, see E. Zeller (ed.), Die Philosophie der Griechen in
ihrer geschichtlichen Entwicklung (Leipzig, 19192246), 3.599; Edelstein (n. 2), 300, n. 58;
Festugire (n. 1), 21; Kidd (n. 2), 1.980; Hoven (n. 2), 62. Unfortunately we do not know the
original context of Posidonius remark. There is, for example, no evidence to support Zellers
conjecture (see above) that it was a commentary on the Phaedrus.
36 Pl. Tim. 30b31a, 36e37a, Leg. 10.8919.
37 Gal. PHP 5.4668, pp. 322.28326.8 De Lacy (fr. 31 EK).
38 Hackforth (n. 34), 645 suggests that the argument of the passage cannot be regarded as a
direct argument for the immortality of individual souls; but it is reasonable to believe and
indeed, since it is the individual soul that Socrates will be concerned with in the myth, we cannot
avoid believing that Plato regarded a demonstration of the immortality of soul in general as
applicable to individual souls.
39 For Hermias reading of all soul as noetic soul, see Robinson (n. 1), 111.
40 Cf. n. 10.

122

A.E. JU

if taken to refer to the whole of soul, appropriate to the world soul in the Stoic sense.
But my preliminary observations tend to suggest that Posidonius interpretation
rather rested on the context in the dialogue, and that by the expression the world soul
he meant the whole of soul as understood by Plato.
In pursuing these points further, we can begin by looking at a passage of the myth
immediately subsequent to Platos argument noted above. In the passage, Plato, using
the precise words all soul, continues:

[].
All soul has the care of that which is soulless, and travels the whole universe, coming to be at
different times in different forms. So, when it is perfect and winged, it journeys in the upper
region, governing the whole world; and when it sheds its wings, it comes to fasten on something
solid, and, settling there, takes to itself an earthy body []. (Pl. Phdr. 246b6c4)

It remains a matter of speculation whether Posidonius referred to the passage of the

myth just quoted, since Hermias is silent on the point. But supposing Posidonius did
so, the inference would be that he found warrant to believe that all soul represents for
Plato the whole of soul, which gives birth to individual souls, as these are in Platos
words coming to be at different times in different forms. Further, this same soul,
perfect and winged, and in particular in Platos words governing the whole world, is
likely to have coincided in Posidonius estimation with the world soul as certainly no
individual soul can do this thereby cementing the equation of all soul with the
world soul.
In seeking to clarify Posidonius interpretation further, we can return to the
argument of Plato quoted earlier. The question whether in the opening line of the
argument Plato used the expression what is self-moving or what is eternally moving
has yet to be resolved by scholars. Leaving aside the debate on this crux, what is clear
in subsequent lines is that Plato, with a distinction between what is self-moving and
what is moved by something else, maintains that only what moves itself does not
cease its motion. He proceeds to prove this self-mover to be the source and the first
principle of motion for all other things that are moved, further identifying soul as
this self-mover. Plato concludes with the syllogism that what is self-moving is
immortal; what is self-moving is soul; therefore soul is immortal.
In speaking thus of soul as the self-mover, Plato offers an interesting disproof of
the possibility both that it should come into being and that it should be destroyed,
on the grounds that the former has already been proved false and the latter is
inconceivable. What he takes to be inconceivable is that the whole heaven and the
whole of becoming, comprising all things that make up any possible universe, should
stop and no longer move, for the reason that no source of motion to bring it back into
being would exist; but this is inconceivable; therefore, the self-mover cannot be
destroyed.
It is likely that Posidonius was specifically engaged by this conditional prediction
of Platos noted above, given that before assigning Posidonius to one group of Platos
interpreters, Hermias explains their interpretation by quoting because he [Plato] said
all and a little later adds that,

STOIC AND P OSIDONIAN THOUGHT ON THE IMMORTALITY OF SOUL

123

Were it otherwise, the whole heaven and the whole of becoming would collapse and stop.41

It seems possible that Hermias here is quoting Posidonius own citation of this statement from Platos dialogue. We could, if so, be confident in inferring that Posidonius
placed as much weight on the statement as on the word all; and, given that the
immediate context of the argument including the statement is Platos proof of soul as
a self-moving substance, it seems possible also to suppose that Posidonius attended
particularly to this Platonic conception of soul.
The question immediately arises why Posidonius placed so much weight on Platos
conditional prediction. Posidonius presumably would not quarrel with the syllogism
as noted above which stands, even if soul applies to an individual soul, such as
Socrates soul. But in my view, Posidonius judged further that, as far as the
conditional prediction is concerned, Plato was treating the self-mover as a term
corresponding to all soul. That is, the Platonic conditional prediction in Posidonius
eyes stands if and only if the self-mover carries the same referent as all soul by
referring to all the self-moving substance there is. Posidonius formulation makes
sense, in so far as it is clearly possible to go on thinking of the world in the event of
the destruction of an individual soul, for instance, Socrates soul, while, were we to
suppose the loss or destruction of all the self-moving substance there is, then we might
quite possibly find the motion of the world unimaginable; therefore, this substance
cannot be destroyed.
The above discussion has three major suggestions. First, Posidonius did not neglect
the fact that Platos argument is meant to apply to individual souls. Second,
Posidonius gave attention to the fact that at the opening of the argument Plato said
all, further stressing the Platonic conditional prediction that, should the self-mover
hypothetically be destroyed, the whole world and the whole of becoming would stop
and no longer move. Finally, Posidonius concluded that the truth of Platos
proposition hinged on an identity between all soul and all the self-moving substance
there is. It is likely that this conclusion guided Posidonius in inferring that all soul
must for Plato refer to soul collectively, taken here as equivalent to the soul, in
Platos words, governing the whole world.
Let me conclude with a remark on Posidonius interpretation of Plato discussed
above. It is typical of Posidonius that he read previous philosophers and especially
Plato constructively for his own purposes, rarely venturing purely disinterested
comments. Taking this characteristic into account, it seems possible that Posidonius
used his reading of Plato to support, or at least refine, an essentially Stoic view of
self-moving soul.42
Relevant sources for Posidonius reflection on Platos idea of self-moving soul are
Achilles Tatius, Intr. in Arat.13 and perhaps also Sextus Empiricus, Math. 9.712.43
41 I am grateful for, and accept, the suggestions of A. Mourelatos (in my talk in the University
of Texas at Austin) that the phrase quoted above, rather than referring to the universes annihilation in its entirety, may imply that it suddenly stops and does not move any longer, as if it and
its occupants remain entirely frozen.
42 Macr. In Somn. Scip. 1.14.19 (fr. 140 EK) presents a list of the philosophers Philolaus, Plato,
Xenocrates and Aristotle, who are said to have viewed soul as a self-moving substance; this
doxography is plausibly traced back to Posidonian material.
43 Cf. n. 6. Kidd and Edelstein, in their Fragments, include only the Achilles passage, whereas
J. Dillon, The Middle Platonists (London NY, 1977), 11012 takes Sextus, Math. 9.712 as
typically Posidonian. Kidd (n. 2), 2.550 also remarks on this Sextus passage as closer still to the
Posidonian expression of the concept as presented in the Achilles passage. In the main I agree, on

124

A.E. JU

Any closer scrutiny of these passages notwithstanding, it seems evident that the
passages sources equally insist, against Epicureanism, on the existence of souls in the
extra-terrestrial region, whether these souls be surviving human souls or daemons or
the souls of the stars; and that, analogically or otherwise, both bear out a Posidonian
idea of the souls causal action upon itself through its capability of holding itself
together. In view of this Posidonian preoccupation, it remains possible that when
Posidonius interpreted Phaedrus 245c, his concern lay with the broader context of
dealing with the topic of self-moving soul. Possibly, in engaging his contemporaries
the Epicureans, Posidonius both had recourse to his Stoic predecessors and solicited
the support of Plato by referring to previous conceptions of self-moving soul.
Other material for (perhaps mainstream) Stoic reflection on the Platonic idea of
self-moving soul can be found at Sextus, Math. 9.76.44 Leaving aside the question of
Sextus Stoic source, there seems little room for doubting that the passages argument
represents a standard Stoic doctrine, to which Posidonius seems to have taken no
significant exception. The passage itself is roughly reminiscent of the Phaedrus
passage that we have been studying, in two basic respects, namely self-motion and
eternal motion, despite differences in the detail of the passages argumentation. From
a comparison of the two, it appears that the source of the Sextus passage took over
from Plato the distinction between what moves itself and what is moved by something else. In the first part of the argument, denial of a self-moving power is accused
of the absurdity of an infinite regress. The author proceeds to prove this self-moving
power to be divine and everlasting, maintaining the impossibility of its being in
motion from some point in time, as there could then be no cause of its motion;
therefore it must be in motion from eternity. The general ideas of the argument and
in particular the term self-moving are strong confirmation that the Stoics
conception of such a self-moving power, taken here as equivalent to god or Stoic
reason, echoed the Phaedrus passage discussed above even when they were not
interpreting it directly.45
Seoul National University

A.E. JU
julepli@hotmail.com

the basis that, although the Sextus passage does not refer to Posidonius by name, it would seem
baseless to deny its similarity with the Achilles passage, both doctrinally and in textual features.
But, for lack of direct evidence, it appears difficult to determine that the subsequent section of
Sextus, Math. 9.736 derives from a Posidonian exposition. However, the arguments of this
section seem to represent standard Stoic doctrines, to which Posidonius seems to have taken no
significant exception.
44 Cicero, at Tusc. 1.535, discussing previous philosophers views on the immortality of soul,
quotes the same argument of the Phaedrus to which Posidonius referred, except for the opening
sentence
. Cicero possibly knew Posidonius interpretation of the
argument. Judging from Ciceros translation of
as quod semper
movetur aeternum est (1.53), and his accompanying comment (1.55), he favoured
.
Sextus, Math. 9.76, on the other hand, tends to suggest that Sextus Stoic source employed
. It seems possible that Cicero differed from the Stoics in interpreting Phdr. 245ce;
yet it is unclear whether their difference was associated with the competing positions of the Stoic
and Academic schools over this crux.
45 This article has benefited from discussion with Christopher Gill, Alexander Mourelatos,
Terumasa Ohkusa, Malcolm Schofield, David Sedley, Stephen White and Hoyoung Yang. I thank
the editors and anonymous referee of CQ for suggestions and corrections.

Classical Quarterly 59.1 125131 (2009) Printed in Great Britain

doi:10.1017/S00098388090000093

125
HORACES PRIAPUS: A LIFELOWELL
ON THE ESQUILINE
EDMUNDS

HORACES PRIAPUS: A LIFE ON THE ESQUILINE

(SAT. 1.8)*
Horaces Priapus has traits familiar from the poems of the Priapea.1 These traits are
self-consciousness concerning ligneous origin, red phallus, raised arm, and protection
of gardens against thieves.2 Yet he has one trait unparalleled in the Priapea, a reed
attached to his head. The reed scares birds away, and this scarecrow function is presumably also generic, whether or not there was always a reed.3 Horaces uolucres /
uetat nouis considere in hortis (67) seems to refer to this particular function,
because scarecrows protect newly sown gardens, where birds eat the unsprouted
seeds.4 Does Priapus speak of himself, then, as a new installation, and is the reader to
assume that he has been fitted out as a scarecrow because these gardens are new? So
one might at first assume, but already in line 8 Priapus begins to tell the history of the
place. (It was once, he says, a potters field, marked by a cippus that defined its
* I am grateful to Chrystina Huber and John Bodel for corresponding with me and for their
great generosity in sending me copies of their writings; to Niklas Holzberg for corrections, suggestions and most of the references in n. 32; to Denis Feeney for information on the phenomenon
discussed in n. 23; to the anonymous referee and Rhiannon Ash for helpful comments.
1 In this article, the title Priapea (abbreviation Priap.) refers to what can be considered the
standard collection of eighty epigrams about Priapus. On the collection and its various contents
and names see Note to the reader in W.H. Parker, Priapea: Poems for a Phallic God (London,
1988), facing p. 1. The date of this collection is controversial: see C. Goldberg, Carmina Priapea:
Einleitung, bersetzung, Interpretation und Kommentar (Heidelberg, 1992), 356. Even if it is
entirely later than Horace, the epigrams have comparative value for Sat. 1.8 because their motifs
would have been known to Horace. So Priapic poems by Catullus suggest: 47.4 uerpus Priapus
ille (Phalaecian); fr. 1.1, 2 (Priapean). Cf. 16.1 pedicabo uos et irrumabo with the same pair of
threats in Priap. 35.12, 5. J. Uden, Impersonating Priapus, AJP 128 (2007), 126 studies poems
16, 47, 56 as Catullus experiments in the genre of Priapic poetry. For that matter the phallic
threats appear already in Leonidas of Tarentum (A. Pl. 236 = lxxxiii GP; A. Pl. 261 = lxxxiv
GP). Consider also the consistency of the representation of Priapus in art in the period 100
B.C.E. to 200 C.E.: W.-R. Megow, Priapos, in J. Boardman et al. (edd.), Lexicon Iconographicum
Mythologiae Classicae 8.1 (Zurich, 1997), 1044.
2 Selfconscious: Priap. 6.1, 10.4, 25.1, 43.1, 56.3, 63.912, 73.3; [Virg.] Priapeum 2.1
Richmond (OCT) = 85.1 Bcheler, etc. Red: Priap. 1.5, 26.9, 72.2, 82.8; Tib. 1.1.17; [Tib.]
Priapeum 2.8 Luck = 82 Bcheler = Priapeum Quid hoc noui est? Richmond (OCT). Raised arm,
with club or falx (weapon or implement not specified in Horace): Priap. 6.2, 11.2, 20.5, 30.1, 33.6,
55.1.
3 This function of a Priapus elsewhere only at Tib. 1.1.18, where it is not a reed but the falx:
terreat ut saeua falce Priapus sua; Virg. G. 4.11011; Ov. Fast. 1.400, in the much discussed story
of Priapus failed rape of Lotis, where it is the phallus, quique ruber pauidas inguine terret aues.
Despite the impression conveyed by H. Herter, De Priapo (Giessen, 1932), 199, Cornutus Theol.
Graec. says nothing about Priapus protecting gardens against birds. Cf. Lactant. Div. inst. 2.4.2:
fures enim tam stulti sunt ut Priapi tentiginem timeant, cum aues ipsae, quas terrore falcis aut
inguinis abigi existimant, simulacris fabre factis id est hominum plane similibus insidant nidificent
inquinent (cited by Herter, op. cit., 212). A reed is not mentioned, and Herter exaggerates when he
says harundo saepe in capite eius fixa erat. The short list of representations of Priapus in
various media which he gives (200) hardly bears out this statement.
4 See the second of Porphyrios comments on line 7, quoted in n. 7 below. Note that Porphyrio
was aware of the ambiguity of nouis. In an established garden, birds might have been pleasant
companions of Maecenas. Cf. Pieridas Phoebumque colens in mollibus hortis / sederat argutas
garrulus inter aues (Eleg. Maec. 1.356).

126

LOWELL EDMUNDS

dimensions [813].) One assumes that he is speaking from his own memory and must
pre-date the new gardens. What, then, had he been doing there? Was he threatening
thieves with anal penetration, as his subsequent reference to thieves might suggest
(1718)? Was he a scarecrow?
The answer to these questions, and to other questions soon to be posed, will
depend upon the readers ability to fill in the blanks in this Priapus elliptical life story.
It will be suggested that the best way to do so is to assume that Horaces persona
loquens was a real Priapus, i.e. a physical image which the first readers or auditors of
the poem could see. Further, as a perceptible physical entity, he had a particular
location on the Esquiline, known to the first audience. He speaks, like his generic
cousins, from a fixed deictic centre, indicated at the outset by huc (8).5
Even if Horaces Priapus is not real but notional, as many or most of the Priapuses
in the Priapea are notional (i.e. the epigrams are not inscriptions in a shrine or
somehow posted near the figure of a Priapus, despite their conceit of realism), the
question of location will not go away. The modern reader is not spared the task of
coordinating Horaces Priapus words with his location; cannot otherwise make sense
of Priapus as the speaker of this poem; a fortiori, has no access to the mind of Horace
except on Priapus terrain.
To return to Priapus account, the Esquiline, that is the part of the Esquiline
rendered foul by the cemetery, has become habitable, and one can take walks along
the agger (1416), that is south of the Esquiline Gate.6 As readers of Horace have
known since the time of Porphyrio, if indeed it was ever forgotten, Priapus is talking
about Maecenas clean-up of the area and his building of his own estate.7 But it would
be odd if a Priapus had been guarding a cemetery like the one this Priapus describes.
What would he have guarded? Why would the thieves, with whom Priapus was once
concerned (1718), have come into such a place?
As for me, Priapus goes on, my role has changed, along with the changes in the
area. I am less preoccupied by thieves and wild creatures than by witches, who come
5 For the concept of deictic centre, I cite the foundational work, K. Bhler, Sprachtheorie:
Die Darstellungsfunktion der Sprache (Jena, 1934), 10220 (Die Origo des Zeigfeldes und ihre
Markierung). Centre is now commonly used instead of Bhlers Origo. As a statue or
statuette, and the persona loquens of most of the epigrams, Priapus is the perfect deictic centre
or deictic zero point. For huc, cf. Priap. 12.5, 14.1, 63.9, 64.2, 69.2, 70.5, 77.3, 80.10. hic occurs
12 times and hinc 5 times. The combined total of the occurrences of the three deictic adverbs is
26.
6 J. Bodel, Graveyards and Groves: A Study of the Lex Lucerina, American Journal of Ancient
History 11 (1986 [1994]), 52; id., Dealing with the dead: undertakers, executioners and potters
fields in Ancient Rome, in V.M. Hope and E. Marshall (edd.), Death and Disease in the Ancient
City (London, 2000), 1312; C. Huber, Zur Topographie der Horti Maecenatis und der Horti
Lamiani auf dem Esquilin in Rom, Klner Jahrb. Vor- u. Frhgesch. 23 (1990), 11107 (best
maps, loose-leaf, by Helga Stcker); ead., Horti Maecenatis, in E.M. Steinby (ed.), Lexicon
Topographicum Urbis Romae 3 (Rome, 1996); T.P. Wiseman, A stroll on the rampart, in M. Cima
and E. La Rocca (edd.), Horti Romani (Rome, 1998), 13 For maps see also Huber, Das
Archologische Informationssystem AIS ROMA: Esquilin, Caelius, Capitolium, Velabrum,
Porta Triumphalis, Bullettino della Commissione Archeologica Comunale di Roma 106 (2005
[2006]), 959. It is not clear why at Sat. 2.6.323 Horace says atras / Esquilias. See the
questions raised by A. Kieling and R. Heinze, Quintus Horatius Flaccus: Satiren (Berlin, 1957),
ad loc.
7 Porphyrio (Hauthal) on line 7: (1) Ideo dixit, quod, cum Esquilina regio prius sepulchris et
bustis uacaret, primus Maecenas [ad] salubritatem aeris ibi passus hortos constituit. (2) Potest nouis
hortis accipi pro recens satis. Maxime enim aues tum prohibendae ex hortis sunt, ne semina in terra
missa colligant. I have numbered the two parts of the gloss. Cf. Hor. C. 3.29.512 for Maecenas
estate.

HORACES PRIAPUS: A LIFE ON THE ESQUILINE

127

here to collect bones and noxious herbs.8 Though he says that he cannot find a way to
keep them out, he proceeds to tell a story about how he once drove a pair of these
intruders away, and this story takes up the rest of the Satire. They had come into the
noui horti by night and were performing their abominable rites, when Priapus farted so
loudly that he split his wooden buttocks and sent them running off in confusion
(4650).9
So concludes the story and so also concludes the poem. As a conclusion to the
poem, the split buttocks are most effective if they correspond to an observable peculiarity of a real Priapus, one which Maecenas and the first readers or auditors of this
poem could see. So Adolf Kieling long ago suggested.10 In this way the story about
the witches becomes a surprise aition for a physical feature which a modern reader can
easily imagine. In his years of tenure on the Esquiline, Priapus had dried out, and the
wood had split along the crease between the buttocks.11 Whether or not Kielings
suggestion adds a virtual image to the collection of artefacts from the Horti of
Maecenas which one now sees in the Museo dei Conservatori in Rome, it certainly
gives the modern reader a useful way of thinking about Horaces Priapus as the
speaker of the poem.
With an old, cracked Priapus in mind, one can try to give a more definite answer to
the question of Priapus relation to the new gardens. Assuming that Priapus location
has remained unchanged, larger new gardens, that is the urban villa, must have
absorbed the garden in which Priapus began his life. The question then becomes: what
garden was it? Pierre Grimal, who was aware of the difficulties of the phrase new
gardens, thought that they were an extension of older gardens of Maecenas, in which
Horaces Priapus would already have had a place.12 This suggestion would solve the
problem, and is on the right track, but there is no evidence for older gardens of
Maecenas.13
A more likely possibility is a pre-existing tomb garden or kepotaphion somewhere
in the vicinity of the potters field. Two inscriptions are cited for the role of a Priapus
in such a setting. One is from Verona: after Dis Manibus C H C there is added locus
adsignatus monimento in quo est aedic(u)la Priapi (CIL 5.3634). It sounds as if the
shrine of Priapus was already there and is mentioned as a way of defining the area of
the tomb. The other inscription is a couplet in iambic senarii: custos sepulcri pene

8 Lines 1722. Note that thieves and birds (3) have become furesque feraeque suetae (17).
Birds are already forgotten as objects of his attention. They return, in the form of crows, and he is
now the possible object of their attention: mentior at si quid, merdis caput inquiner albis /
coruorum (378).
9 J. Hallett, Pepedi / diffisa nate ficus: Priapic revenge in Horace, Satires I.8, RhM 124 (1981),
3417 explores anal associations of the word ficus (Horaces Priapus is made of fig-wood) and
suggests that, because of these associations, this poem redefines the typical (threatened) Priapic
punishment, anal penetration. Here the anus is not the object of Priapic punishment but its
instrument. If this interpretation is possible, it would not be inconsonant with the point I am
making about Priapus loss of generic identity in this Satire.
10 Kieling and Heinze (n. 6), 136, Anla zu den lustigen Versen (i.e. Sat. 1.8) wird gegeben
haben, da eines der Priapbilder in Maecenas Grten wirklich die Beschdigung aufwies, deren
Entstehung hier erklrt wird. I refer to Kieling without having checked pre-Heinze editions.
Heinze says in his preface to the fifth edition (1921), ganz unverndert ist kaum eine Seite
geblieben.
11 Cf. Priap. 48 for observable physical peculiarity; perhaps also 26.
12 P. Grimal, Les Jardins romains (Paris, 19692), 1435.
13 So Huber (e-mail message to me 3 Apr. 2007).

128

LOWELL EDMUNDS

destricto deus / Priapus ego sum. mortis et uitae locus.14 This guarding of a tomb is
unique in Priapic texts (and this metre is attested in a Priapic poem elsewhere only in
[Tib.] Priapeum 2 Luck = Priap. 83 Bcheler). These two inscriptions are the first two
items in Hans Herters inventory under the heading De Priapo mortis et uitae deo.15
But in this inventory one will not find, after the inscription just quoted (custos
sepulcri, etc.), any certain example of a Priapus guarding a tomb. One has to distinguish, then, between guardianship of tomb and of garden belonging to tomb.
There were many such gardens, many kepotaphia, on the outskirts of Rome.16 They
are described in detail in inscriptions, for example, to this tomb belong the vegetable
garden which is enclosed within the wall, and the summer house built beside the door
to serve as a porters lodge (CIL 6.13823) or shops, three in number, to the left and
right of the tomb and the enclosed market garden within (CIL 6.31852).17 One
would expect to find a Priapus in such gardens, which were, after all, his normal
venue.18
These vegetable and market gardens, even with no connection to a kepotaphion,
should in fact be considered. They could certainly be found on the Esquiline outside
the agger, as elsewhere on the outskirts of Rome,19 and Maecenas project, which he
began in c. 40 B.C.E., was easily large enough to include them. It took in an area far
more extensive than the potters field, already large by todays standards at 1,000 by
300 feet (Sat. 1.8.12), or almost seven acres.20
Cicero in Pro Cluentio (66 B.C.E.) preserves a picture of the place from about a
quarter of a century earlier. It appears in a catalogue of the murders perpetrated by
Oppianicus, the father of the plaintiff. One of his victims was Asuvius, a rich young
man whose fortune he hoped to acquire. He lured him from Larinum to Rome. There,
Oppianicus henchman, Avillius, forged a will, signing it with Asuvius name and
making Oppianicus the heir. Asuvius , quasi in hortulos iret, in harenarias quasdam
extra portam Esquilinam perductus occiditur (37). Asuvius, gullible though he was,
would not have gone outside the Esquiline Gate quasi in hortulos unless he knew or
could have believed that hortuli were in fact to be found in this place.
14 CIL 6.3708 = 5173 = ILS 3585 = CLE 193 = 153 Courtney. Courtney comments, Priapus
guards a tomb placed in a garden or vineyard. But the report of the find-spot given in CIL is
only near the monument, i.e. the one from which 48815172 come and for which a plan is given
on p. 926, along with quotations from Giovanni Campanas excavation report.
15 Herter (n. 3), 22932. Megow (n. 1), 1029 states that the protective function of Priapus
extended to houses and graves, citing for the latter Herter (n. 3) and a museum catalogue of
Gallo-Roman figurines, M. Rouvier-Jeanlin, Les Figurines gallo-romaines en terre cuite au Muse
des antiquits nationales (Paris, 1972).
16 For this phenomenon, see J. Bodel, Roman Tomb Gardens, forthcoming in W. Jashemski
(ed.), Gardens of the Roman Empire (vel sim.) (Cambridge).
17 The translations are those of N. Purcell, Town in country and country in town, in E.B.
MacDougall (ed.), Ancient Roman Villa Gardens (Washington, DC, 1984), 188, n. 4.
18 But not in a cemetery, despite P. Lejay, Oeuvres dHorace (Satires) (Paris, 1911), 217: Un
Priape ntait pas dplac dans un cimitire.
19 Suetonius called the Pincian Hill the collis hortulorum (Ner. 50).
20 An acre is 43,560 square feet. If Huber is right about the location of the grove of the
Querquetulanae uirae (cf. n. 25 below), then Maecenas gardens extended as far south as the
modern Via Labicana. She believes the so-called horti Lamiani were part of the gardens of
Maecenas (see Il luogo del ritrovamento del gruppo del Laocoonte e la domus Titi imperatoris
(Plin. Nat. Hist. 36,3738), in Laocoonte alle Origini dei Musei Vaticani [Rome, 2006], 417) and
estimates the size of the gardens as about 62 acres (25 hectares), while the size would have been
about 35 acres (14 hectares) if, as others believe, the gardens of Maecenas and the so-called horti
Lamiani were divided by the former ancient Via Merulana (e-mail message to me 23 Apr. 2007).

HORACES PRIAPUS: A LIFE ON THE ESQUILINE

129

Livys description, applying to a much earlier time (211 B.C.E.), coincides in two
details with what is known of the Esquiline at the end of the Republic and the
beginning of the Principate. Hannibal is only a few miles from Rome. The consuls
organize the defence of the city. Amongst other things they decide to send a band of
1,200 Numidian deserters from their bivouac on the Aventine to the Esquiline: media
urbe transire Esquilias iusserunt, nullos aptiores inter conualles tectaque hortorum et
sepulcra et cauas undique uias ad pugnandum futuros rati (26.10.56).21 The gardens
will reappear in Ciceros description (though Cicero refers to hortuli and perhaps
means a different kind of garden).22 The tombs will reappear in the Satire under discussion. Priapus can see magna sepulchra.23
Renewal of the Esquiline outside the agger had to deal with a variegated terrain,
put to various uses over the centuries.24 From Cicero one knows also of the sandpits
in which Asuvius met his end. If there was an oak tree tall enough to provide shade
for Maecenas, it had to have preceded the horti, as Huber has argued, proposing
that it belonged originally to a sacred grove which she identifies as that of the
Querquetulanae uirae.25 It has been suggested that the kolumbethra constructed by
Maecenas (Dio Cass. 55.7) was a thermal spring found on the spot and reused by him
for his own purposes.26 Similarly, the old Priapus of Satire 8 could have been found in
one of the gardens referred to by Cicero and left there, somewhere within eyeshot of
the potters field.

21 Cf. Tacitus description of a place not far from the city, to the north, on the Via Salaria,
where Petilius Cerealis, sent ahead by Antonius with a thousand horse, was routed by a motley
band of Vitellians: inter aedificia hortosque et anfractus uiarum (Hist. 3.79). See K. Wellesley,
Cornelius Tacitus: The Histories, Book III (Sydney, 1972) on anfractus uiarum: the north-eastern
suburbs of Rome beyond the Castra Praetoria seem to have contained a number of cemeteries
and suburban estates, connected with the main roads by rough tracks; cf. Suet. Nero 4849.
22 These gardens make one think of the garden plots which one still sees in cities in the U.S.,
England and Europe: the English allotments, the German Einzelgrten in a Kleingartenanlage, the Italian orti urbani, the French jardins potagers communautaires or potagers
urbains municipaux. Tecta makes one think of the charming, ingeniously constructed sheds that
these gardens sometimes have.
23 Line 36. Cf. Cic. Ph. 9.17 for such a tomb on the Esquiline. See J. Bodel, Monumental
Villas and Villa Monuments, JRA 10 (1997), 201 for monumental tombs on villa properties.
24 See Bodel (n. 16) on the competition for space outside the walls: Romans looking for extramural burial sites competed for space not only with the suburban villas of the wealthy but also
with warehouses, manufactories, transportation depots, markets, and squatter communities of
laborers (notably teamsters, tanners, and brickmakers) and marginalized groups such as funerary
tradesmen, foreign immigrants, and devotees of certain exotic cults. The best-known case is
Ciceros search for horti for a shrine for Tullia, an obsessive theme of Att. 1213 (4544 B.C.E.). As
he himself says, de hortis etiam atque etiam rogo (12.22.3). See also N. Purcells evocation of what
a person would have seen as he or she left Rome through the Esquiline Gate in 55 C.E.: (n. 17),
1878 and, more specifically on the area of Maecenas estate, Huber (n. 6 [1990]), 1067 with fig.
73. Inside the wall, in the Vicus Sabuci (see Huber 1990, loose-leaf Map 2 E6), where kilns have
been found, Maecenas seems to have taken over a business district as part of his urban renewal:
see Huber, op. cit., 106.
25 maluit umbrosam quercum (Eleg. Maec. 1.33). Huber, Wald und Siedlung im antiken Rom
Spuren heiliger Haine auf dem Mons Oppius, Siedlungsforschung: Archologie Geschichte
Geographie 19 (2001), 59, 76 (with citation of F. Castagnoli, Il tempio romano: Questioni di
tipologia e di terminologia, PBSR 52 [1984], 20, n. 85 for the incorporation of pre-existing
shrines in private estates), 88; for the location Huber (n. 6 [2005(2006)]), 16, 18; (n. 20), 45.
26 S. Rizzo, Horti Mecenatiani, in 7c. Gli Horti dellEsquilino, in Roma Capitale
18701911, 7: L Archeologia in Roma capitale tra sterro e scavo (Venice, 1983), 1956.

130

LOWELL EDMUNDS

One should consider, then, the possibility that Priapus is punning when he says
nouis in hortis, simultaneously using horti in two senses: as the plural of hortus
garden and as the plural horti in the sense of urban villa (cf. OLD s.v. hortus 2).27
Priapus life story begins with a description of his generic function in gardens in the
primary sense, including two lines on his scarecrow function in new gardens, that is
newly planted gardens. He goes on to say, in effect, that he has outlived his generic
function and, in the setting of new gardens, that is Maecenas new estate, now has a
new, post-generic function to guard against witches.
The generic self-description with which the Satire opens would, then, refer first of
all to Priapus functions in that old garden (hortus in the primary sense), which would
have been either a tomb garden or a market garden or a family vegetable garden. He
would have continued to play these generic roles in the indeterminate period
extending from the inauguration of Maecenas new gardens (horti in the sense of
urban villa) up to the moment in which Priapus is speaking (nunc, 14). But in this
period he has gradually (as non tantum, 17 implies) shed these roles, though the main
traits of his physical identity, phallus, raised arm and reed, remain the same. The
ambiguity of nouis in hortis (7) covers this diachronic range from the time of
someones old garden with newly planted seeds to Maecenas incorporation of this
old garden into his vaster gardens from someones market garden or family
vegetable garden or kepotaphion to Maecenas pleasure gardens. The present tenses of
terret and of uetat (7) are also ambiguous, complementing the ambiguity of new.
After deus inde ego (sc. eram or sum?), these present tenses sound at first like an
imperfective, descriptive present, that is, descriptive of what Priapus is actually doing,
but they turn out to be a general present expressing a vrit dexprience concerning
Priapuses which no longer applies to this Priapus existence.28
An antique, fissured Priapus, one losing his physical integrity, a Priapus physically
recontextualized in new horti in which he abandons his traditional functions he
would be the perfect image for loss of poeticgeneric identity. In his new role Priapus
is post-generic in this sense too. He makes fun of himself as a Priapus.29 He is not
speaking, after lines 17, to a reader imagined as reading Priapic verses posted in a
shrine of Priapus or near a Priapus statue. He no longer conveys his characteristic
phallic threats. He has forgotten the metres of the Priapea.30 Speaking in dactylic
hexameters, he refers to those neer-do-wells Pantolabus and Nomentanus (11; cf.
Sat. 2.1.21); he describes Horaces horrid fascination, Canidia, and her companion

27 The suggestion concerning horti as a pun came to me from John Bodel (e-mail message 31
Mar. 2007).
28 On the Latin present see A. Ernout and F. Thomas, Syntaxe Latine (Paris, 1953), 2201.
29 Parody is the theme of the short discussion of Sat. 1.8 by V. Buchheit, Studien zum Corpus
Priapeorum, Zetemata Monographien zur klassischen Altertumswissenschaft 28 (Munich, 1962),
634.
30 Cf. n. 1 above for the comparative value of the Priapea.
31 For Canidia, Hor. Epod. 5, 17; cf. 3.8; Sat. 2.1.48, 8.85. For Sagana, Epod. 5.25. For a survey
of views on the identity of Canidia see D. Mankin, Horace: Epodes (Cambridge, 1995), 299301.
L. Watson, A Commentary on Horaces Epodes (Oxford, 2003), 1889 points out a previously
unnoticed parallel between the boys curses in Epod. 5, introduced by dubius unde rumperet
silentium (85) and Priapus fart in Sat. 1.8: both, by breaking ritual silence, have the effect of
vitiating the witches magical ceremonies.
32 The description of the trio is from P.M. Brown, Horace: Satires I. (Warminster, 1993), ad

HORACES PRIAPUS: A LIFE ON THE ESQUILINE

131

Sagana31; and he refers to a trio of unidentifiable undesirables (39), all in a way that
might remind one of a Satire by Horace.32
Rutgers University
lowedmunds@gmail.com

LOWELL EDMUNDS

loc. Interpretation of this Satire has often called attention to similarities between Priapus and
Horace (i.e. in the persona of satirist): W.S. Anderson, The form, purpose, and position of
Horaces Satire I, 8, AJP 94 (1972), 45; J.E.G. Zetzel, Horaces Liber Sermonum: The
Structure of Ambiguity, Arethusa 13 (1980), 61, 66; J. Henderson, Satire writes Woman:
Gendersong, PCPhS n.s. 35 (1989), 602; this piece was trimmed and rewritten (v) under the
same title in id., Writing Down Rome: Satire, Comedy, and Other Offences in Latin Poetry
(Oxford, 1999), 18891; cf., for the same take on Sat. 1.8, id., Not Women in Roman Satire but
When Satire writes Woman , in S. Braund (ed.), Satire and Society in Ancient Rome (Exeter,
1989), 18891; S. Braund, Roman Verse Satire, Greece & Rome New Surveys in the Classics 23,
(Oxford, 1992), 21; B. Hill, Horace, Satires 1.8: Whence the witches? Thematic unity within the
satire and within the Satires of Book I, in M. DeForest (ed.), Womans Power, Mans Game:
Essays on Classical Antiquity in Honor of Joy K. King (Wauconda, IL, 1993), 261; M. Habash,
Priapus: Horace in disguise? CJ 94 (1999), 2856, 2889, 2956; F. Felgentreu, Horaz, Satiren
I,8 und die Vielfalt der Einfalt, Hyperboreus 5 (1999), 281; T.S. Welch, Est locus uni cuique suus:
City and Status in Horaces Satires 1.8 and 1.9, CA 20 (2001), 1847; E. Gowers, Blind eyes and
cut throats: Amnesia and silence in Horace Satires 1.7, CP 97 (2002), 15960; S. Sharland,
Priapus magic marker, AClass 46 (2003), 105.

Classical Quarterly 59.1 132141 (2009) Printed in Great Britain

doi:10.1017/S0009838809000010X

132
THE ONE AND ONL
Y FO NYN
S BAN
D U SIAE
LLEWEL
MORGAN

THE ONE AND ONLY FONS BANDUSIAE

Nobody knows where Bandusia was, but it is a fair guess that, like the pine tree
Horace dedicates in 3.22, it was on Horaces Sabine estate.1 David Wests statement of
the likely whereabouts of the fons Bandusiae of Odes 3.13, from the third instalment
of his superb introductory edition of Horaces Odes, may be considered representative
of the broad scholarly consensus on this question from late antiquity (thus pseudoAcro at Odes 3.13.1 and Porphyrio at Epist. 1.16.12, where the spring described is
undoubtedly at his villa) to recent times: Horace is addressing the spring behind his
farmhouse.2 Another, more lucid view momentarily prevailed in the latter part of the
eighteenth century (and reappears sporadically elsewhere), as we shall see. The main
contention of this article is that those scholars who have maintained that the fons
Bandusiae was nowhere near the Sabine farm, a small minority, are absolutely right.
But I shall also be suggesting that interesting implications for our understanding of
the poem addressed to the Bandusian spring follow from clarity as to its geographical
location.
I. FINDING THE FONS BANDUSIAE
The true location of the fons Bandusiae was conclusively established some considerable time ago, in Bertrand Capmartin De Chaupys Dcouverte de la maison de
campagne dHorace.3 A French abb, De Chaupy had stumbled by sheer chance on a
bull of Pope Pascal II, dating to 1103, which made reference in passing to Ecclesiam
sanctorum martyrum Gervasii et Protasii in Bandusino fonte apud Venusiam, the
church of the Holy Martyrs Gervasius and Protasius at the Bandusian Spring at
Venusia.4 Presented with such seemingly powerful evidence, De Chaupy decided to
travel to the vicinity himself, and managed to narrow down the location of the church
and spring to the modern town of Palazzo S. Gervasio, which he placed about six
miles (though it is in fact closer to ten) to the east of Venusia (modern Venosa) along
the Appian Way.5 What makes this information compelling, of course, and the link to
* This discussion of a poetic expression of indebtedness owes its own debt to readers and
interlocutors who have indulged the authors determination to pursue his (no doubt eccentric)
view of Odes 3.13. That includes Bob Cowan, Lindsay Watson, Matthew Leigh and the reader for
CQ, and the participants at a symposium on Horace held in honour of Margaret Hubbard at St
Annes College, Oxford in May 2008: especially Gail Trimble, Fiachra Mac Gorain, and the
honorand on that occasion, to whom this article is respectfully dedicated.
1 D.A. West, Horace, Odes III: Dulce Periculum (Oxford, 2002), 120.
2 G. Williams, Tradition and Originality in Roman Poetry (Oxford, 1968), 149.
3 (Rome, 17679), 3.3635 and 53641.
4 The full text may be found at P. Jaff, Regesta Pontificum2 (Leipzig, 18815), no. 5945; the
relevant portion of the bull is also printed at B.D. Frischer and I.G. Brown (edd.), Allan Ramsay
and the Search for Horaces Villa (Aldershot, 2001), 127.
5 De Chaupy (n. 3), 538: six milles au dessus de Venose au lieu appell Palazzo. For the
route of the Via Appia past Palazzo S. Gervasio see P. Vinson, Ancient roads between Venosa
and Gravina, PBSR 40 (1972), 5890, at 68: the Fontana Rotta marked on Vinsons map
directly alongside the line of the road was what De Chaupy identified as the residue of the fons
Bandusiae. The 6 miles S. of Venusia of E.H. Bunbury (n. 14) was evidently a misreading (or
perhaps mishearing from dictation) of De Chaupys au dessus de as au sud de.

THE ONE AND ONLY FONS BANDUSIAE

133

Horaces poem indisputable, is that Venusia was Horaces hometown, and indeed the
papal document also mentions two localities neighbouring the spring, Bantium and
Ac(h)erentia,6 which feature in the (beautifully contoured)7 passage from an earlier
ode in Book 3 where Horace reminisces about his charmed childhood (3.4.1316):
mirum quod foret omnibus,
quicumque celsae nidum Acherontiae,
saltusque Bantinos et aruum
pingue tenent humilis Forenti.
which was a marvel to all who live in the nest of high Acherontia and the high clearings of
Bantium and the rich ploughland of low-lying Forentum.8

It thus seems perfectly clear, as De Chaupy concluded, that the fons Bandusiae was,
contrary to the statements of the ancient commentators, une Fontaine, non de la
Campagne dHorace, mais de la Patrie:9 not a feature of Horaces Sabine estate but a
landmark of Horaces youth in the marches of Lucania and Apulia (Hor. Serm.
2.1.349). Not just a local landmark, it is worth adding: by virtue of the position of
the spring directly alongside the Appian Way, this was a location potentially identifiable also by the inhabitants of the city of Rome. Indeed, if we follow Brodersens
analysis of the Roman mental map, not an essentially cartographical conception like
our own, but linear, structured by routes which only register the relative position of
the landmarks situated on them,10 this landmark on the road between Rome and
Brundisium would necessarily be a prominent feature on the Roman map of Italy.11
De Chaupy was read in turn by Allan Ramsay, whose An Enquiry into the Situation
and Circumstances of Horaces Sabine Villa Written during travels through Italy in the
years 1775, 76 and 77 has recently been published for the first time.12 Ramsay accepted
De Chaupys findings without qualification,13 as have a small rump of scholars
since,14 but in mainstream literary scholarship, at least, the history of this scholarly
6 The name of the latter town is also written as Ac(h)eruntia or Ac(h)erontia: their modern
counterparts are Banzi and Acerenza.
7 D.A. West, Horaces poetic technique in the Odes , in C.D.N. Costa, Horace (London,
1973), 2958, at 345.
8 Modern Forenza.
9 De Chaupy (n. 3), 365.
10 K. Brodersen, Terra cognita. Studien zur rmischen Raumerfassung2 (Hildesheim, 2003), 290,
referring to the intermediate space of countries and regions.
11 The location is not mentioned in the Antonine Itinerary, but as it happens that record
appears to be deficient for this section of the Appian Way. See Vinson (n. 5), 867 for the
likelihood that ancient Siluium is modern Gravina, and that the 20 milia passuum stated as the
distance between Venusia and Siluium in the Itinerary, as contrasted with the 35 m.p. of the
Peutinger Table (a better indication of the real distance between Venosa and Gravina), is the
result of the dropping out of a stage between the two towns from the text of the Itinerary. For the
relevant text, see O. Cuntz, Itineraria Romana (Leipzig, 1929), 17.
12 Frischer and Brown (n. 4): on De Chaupy see 12331.
13 Here [scil. in the papal bull] was not only a fountain of the same name with that which had
been in vain sought for, but a fountain of such name and consideration as to serve as a landmark,
and at the same time so linked in the Bull itself with other places with which Horace was known
to have been connected as to leave little or no doubt of its being the same fountain which he had
celebrated in his ode: Frischer and Brown (n. 4), 1245.
14 Thus W. Smith (ed.), A Dictionary of Greek and Roman Geography (London, 1854), s.v.
Bandusiae fons (E.H. Bunbury): Smiths dictionary has recently been republished by I.B. Tauris
(London, 2005), with an introductory essay by C. Stray. Stefania Quilici Gigli in S. Mariotti (ed.),
Enciclopedia oraziana (Florence, 1996), I.5578 offers a judicious, non-committal discussion of

134

LLEWELYN MORGAN

question took a decidedly peculiar turn. For it transpires that quite a number of the
scholars who have held that the fons Bandusiae was a feature of Horaces Sabinum
have not only been aware of De Chaupys research but have also accepted the
Frenchmans conclusions in all essentials. The idea has gained currency that, although
the original fons Bandusiae was indeed near Venusia, Horace gave to a Sabine spring
the name of a famous landmark near his birthplace, as Nisbet and Rudd have
recently put it:15 but the same notion is to be found in KiesslingHeinze (ad loc.) and
in Fraenkel, who traces it back at least as far as the middle of the nineteenth century.16
The belief shared by all these eminent scholars is, in effect, that there were two
Bandusian springs.
William of Ockham would have had something to say about this. But let us wield
the razor on his behalf: Horace never offers any indication that the fons Bandusiae was
at his Sabine farm. That is the fallacious, although perfectly understandable,
assumption introduced by his ancient commentators, aware as they were of a spring
on the Sabine estate, vividly evoked at Epist. 1.16.1214.17 In addition we do in fact
know where it was, because De Chaupy found all the relevant evidence in that papal
bull. The kindest description of the compromise represented by the two-spring
hypothesis is that it is inelegant. But as an account of a poem with a readership it is
positively incoherent. In the absence of any hint from Horace as to the location of the
fons that he is celebrating, the two-spring proponents are obliged to argue that
contemporary readers would have thought of the landmark on the Appian Way near
Venusia, but then arbitrarily dismissed it in favour of an ersatz fons Bandusiae at the
Sabine villa: this is a lot to expect a reader to do without any guidance whatsoever
from the poet.
II. A NATIVE SPRING
There was only ever one fons Bandusiae: I hope that is now self-evident. As for its
location, it lay ten miles or so to the east of Venusia but I claim absolutely no credit
for that discovery myself. What I would like to do on my own account is to build a
larger argument about this poem on that topographical clarification. I believe that
other (often well-recognized) implications of O fons Bandusiae click into rather
precise focus once the geographical question is settled. Rather more ink has been spilt
on Bandusias limpid waters than the kids blood which has been the focus of modern
(and, I believe, ancient)18 disquiet. But one point of agreement, and this time an incontestable one, is that a central concern of this poem is the poets own achievement.
theories regarding the location of the fons, but the illustration at I.285 is asserted to be of Il fons
Bandusiae a Licenza, i.e. on the Sabine estate. The village of Licenza is named after a river of the
same name, and Licenza is a clear corruption of Digentia, on which see n. 17 below.
15 R.G.M. Nisbet and N. Rudd, A Commentary on Horace, Odes Book III (Oxford, 2004), 173.
16 E. Fraenkel, Horace (Oxford, 1957), 203, n. 1, citing C.G. Zumpt in E.F. Wstermanns 1843
re-edition of L.F. Heindorf s edition of Horaces Satires, 17, n. 1. Quilici Gigli (n. 14), 557
attributes the idea to G. Boissier, Nouvelles promenades archologiques (Paris, 1884), 301. A rum
exercise in Quellenforschung, this: the search for the scholar who first lost the Quelle Bandusia.
17 This spring is described by Horace as identical in nomenclature to a riuus (fons etiam riuo
dare nomen idoneus, 12), seemingly the riuus later named as Digentia (Ep. 1.18.104, gelidus
Digentia riuus). I cannot see any space at the Sabine farm for a spring named Bandusia. For a
discussion of the symbolic value of the Digentia in Horaces verse which tackles issues relevant in
broader ways to this article, see J.C. Bramble, Persius and the Programmatic Satire (Cambridge,
1974), 623.
18 Pp. 13740 below.

THE ONE AND ONLY FONS BANDUSIAE

135

In the last stanza of 3.13 Horace famously states that the spring will achieve fame in
his poetry (me dicente, 14) comparable to the springs of Greek poetry (1316):
fies nobilium tu quoque fontium,
me dicente cauis impositam ilicem
saxis, unde loquaces
lymphae desiliunt tuae.
You too will be one of the famous fountains, as I sing of the holm oak overhanging the hollow
rocks, whence your chattering waters leap down.

The Greek springs to whose company Horaces poetry will raise the fons Bandusiae
are such familiar names as Aganippe, Hippocrene, Castalia, Dirce and Arethusa, all
of them conceived as sources of poetic inspiration and figures for poetry itself, and in
all cases associated with Muses. The mode of thinking about the poetic process
underlying this proliferation of inspirational springs is well described by Steiner in
relation to Pindar: [i]nspiration demands that the poet take in some power from
without, and that he carry within him the force of the divine which makes him truly
entheos, god-possessed.19 The idea that a spring should fulfil this role is extremely
familiar in Roman poetry, and that is no doubt primarily due to Callimachus
/
,
the pure and undefiled little stream that trickles from a holy fountain, the best of the
best (H. 2.11112), and his restaging early in Aetia 1 of Hesiods encounter with the
Muses on Mt Helicon at the start of the Theogony, which involved some reference
both to the Hippocrene and seemingly also to the less elevated (in more than one
sense) Aganippe, source of the Permessus (Call. fr. 2 Pf.; fr. 696 Pf.).20 But as
Callimachus debt to Hesiod implies, the idea is older, and typically (or at least
archetypically) the notion of the poetic spring entails some indication of the
provenance of the poet to whom it provides inspiration. Thus Hesiod of Ascra
learned his poetry from the Muses who danced around and cleansed themselves in the
springs of Helicon (Theog. 134), and Propertius can consequently talk of Ascraei
fontes (2.10.256), a Hesiodic level of inspiration to which he cannot as yet aspire,
acquainted so far only with the Permessus. AP 9.64 similarly has the Muses offering
the inspiring water of Helicon to Hesiod, and Hesiod drinking his fill of it before
penning his classic works: Persius spoofs the same idea at Prologue 1.21 Pindar of
Thebes also identifies his inspiration with a spring in a locality of special relevance to
him when he presents his sixth Isthmian, at its conclusion, as a draught of Dirce
/
(745):
, I shall offer them a
drink of Dirces sacred water, which the deep-bosomed daughters / of golden-robed
Memory made to rise by the well-walled gate of Cadmus.
The spring Arethusa functions rather similarly in (post-Theocritean)22 bucolic
poetry. At [Mosch.] Epit. Bion. 77 Bion is said to have drunk from Arethusa as Homer
had from the Hippocrene (cf. 912); at Virg. Ecl. 10.1, Arethusa is asked to vouchsafe
one last Virgilian exercise in Theocritean mode. In each case this Syracusan spring
19

D. Steiner, The Crown of Song: Metaphor in Pindar (London, 1986), 44.

W. Wimmel, Kallimachos in Rom. Die Nachfolge seines apologetischen Dichtens in der
Augusteerzeit (Wiesbaden, 1960), 22250.
21 My thanks to Lindsay Watson for these references.
22 Arethusa is mentioned at Theoc. Id. 1.117 and 16.102, but not as a source of personal
inspiration.
20

136

LLEWELYN MORGAN

marks Theocritus geographical origins at the same time as it represents the external
source of inspiration for Theocritean verse. The capacity of the waters of a poets
homeland to symbolize his ingenium is surely also relevant to Ovids regular allusions
to his well-watered place of origin at Sulmo, amongst the Paeligni (Am. 2.1.1, 16.12,
3.15.11; Tr. 4.10.3; Fast. 4.6856), not just a case of the poet lingering affectionately
over his own well-watered family farm,23 nor even simply a statement of national or
ethnic loyalty, but a suggestion that it was to these origins that Ovid owed his poetic
creativity: an aquosus homeland (and this characteristic of the territory is invariably
foregrounded by Ovid) is figuratively a land that provides to its alumnus a wealth of
poetic inspiration. Comparable again are Virgils references to the river Mincius at
Ecl. 7.1213 and G. 3.1415, on the banks of which (according to the latter text) Virgil
would raise his poetic temple to Caesar,24 and the role of the Camenae, denizens of
the spring from which the Vestals drew their daily water, as patron spirits of early
Roman poetic activity.
An obvious implication of the final stanza of Odes 3.13 is that, just as the fons
Bandusiae will bear comparison with these Greek springs, so Horace will join the
ranks of the great Greek poets. West notes how Horace glides into Greek syntax in
this stanza as he effectively claims membership of the Greek literary club, just to
ensure the point is registered.25 The sentiment is in effect a confident assertion of the
mere aspiration to gain inclusion among the lyrici uates that Horace had expressed at
Odes 1.1.356, but with its gesture towards the poets origins near Venusia we are
closer to the terms of 3.30.1014, where the emphasis is on Horaces ascent, ex humili
potens, from humble beginnings in Apulia to literary celebrity. In 3.13 the assessment
of Horaces origins is more positive, but there must still be a hint of irony in the
assimilation of the fons Bandusiae to the grand springs of Greece, some suggestion of
Horaces achievement in elevating a locality so parochial (for all its proximity to the
longarum ... regina uiarum, Stat. Silv. 2.2.12) to the universal celebrity of its Greek
counterparts: the personified lymphae of the last line, more or less interchangeable
with nymphae,26 call to mind, also with some self-belittling irony, the normal denizens
of poetic springs, Musae.27 At any rate the combination of an acknowledged attempt
on Horaces part to equate the fons Bandusiae with springs symbolic of poetic
inspiration and the very personal associations attaching to this landmark in his
homeland makes it natural to assume that other elements of O fons Bandusiae
contribute to some kind of statement about Horaces poetic achievement that this is
indeed the burden of the poem. With this possibility in view, I turn now to the most
controversial part of Odes 3.13, Horaces sacrifice of the goat kid.

23

E. Fantham, Ovid, Fasti Book IV (Cambridge, 1998), at Fast. 4.6856.

It is no doubt significant that Virgils inspirational watercourse at this transitional moment
in his poetic development is not a Callimachean spring but a river, albeit a generically conflicted
one: on the one hand the ingens Mincius meanders with tardis flexibus (14); on the other it
delicately fringes its banks with tenera harundine. For the contradictions inherent in the
generic self-positioning of this passage see Ll. Morgan, Patterns of Redemption in Virgils
Georgics (Cambridge, 1999), 505; and for the observation that in every programmatic utterance
of the Georgics Virgil characterizes his position as transitional, see R.F. Thomas, Virgil,
Georgics (Cambridge, 1988), I.2.
25 West (n. 1), 120.
26 West (n. 1), 121.
27 On the relation of nymphs to Muses, see R. Coleman, Vergil, Eclogues (Cambridge, 1977) at
Ecl. 7.21.
24

THE ONE AND ONLY FONS BANDUSIAE

137

III. SACRIFICE AND SURROGACY

Yet another widely shared assumption of scholarship on this poem concerns the
context of the offering, the life of a goat kid, that Horace is proposing to make to the
spring. Most commentators have concluded that a ritual offering to a spring must
indicate a known festival, the Fontinalia of 13 October, in which case the dramatic
date of the ode (given cras at 3) would be 12 October. Nisbet and Rudd are rightly
sceptical, although they posit in the Fontinalias place another public festival, the
Neptunalia of 23 July.28 But the poet in fact offers no indication that an official
festival, let alone which festival, is being observed, and the assumption that some such
publicly celebrated ritual must be at issue in the poem is in its way as strange, and as
untrue to any natural reading of the poetic text, as the notion of an imitation fons
Bandusiae at the Sabine estate. If a specific festival is entailed Horace tends to offer
strong clues to that effect: we might think of the Matronalia at Odes 3.8.1, Martiis
caelebs quid agam Kalendis, or the more complex play with dates in Odes 1.31, where
the dramatic scenario shifts meaningfully from 9 October to 11 October 28 B.C.29 We
today have to struggle to identify a festival which would fit Horaces account, and it
does not seem that Horace offers information that would have made it any easier for
his contemporaries; in which case it is reasonable to assume that the poet did not want
us to think in terms of a public ritual: cras at 3 is hardly enough.30 Perhaps we should
take our cue from Horaces failure to give us one.
The details of the offering would certainly seem to support a reading of the
sacrifice as a private ritual on Horaces part. As Nisbet and Rudd report, offerings to
springs could take a number of forms, garlands and also pigs and sheep,31 but this
does not bring us very close to Horaces precisely delineated offering of a prepubescent kid, and thus in no way precludes our looking for a significance and
symbolism in the sacrificial victim of more immediate relevance to the poet. And our
attention is undoubtedly focussed upon the kid: that much is ensured by Horaces
arresting decision to dwell, in seemingly unnecessary detail, on the character, the
potential and then the death of the sacrificial victim (38), thereby evoking the pathos
of the creatures unfulfilled promise in a manner very unlike other, unembellished
references to kid sacrifice at Epod. 10.23 and Carm. 1.4.12. But if we are being asked
to contemplate this victim with unusual sympathy and attentiveness, what is it that he
could represent?
One approach to explaining this anomalous focus on the object, process and
consequences of the sacrifice would be to consider the victim, a goat kid on the cusp
of maturity and a fulfilled life, in the light of a well-recognized ancient understanding
of sacrifice as a process of substitution: the victim is offered in exchange for benefits
28

Nisbet and Rudd (n. 15), 1734.

For an excellent account, see D. West, Horace Odes I: Carpe Diem (Oxford, 1995), 14651.
30 Nisbet and Rudd (n. 15), 173: When Horace says that a kid will be offered to the spring
tomorrow (3), he seems to be thinking of a particular festival. It might be countered that lyric
poetry, stereotypically the product of night-time symposia, is programmed to anticipate the
following day, generally with a view to dismissing it in favour of the pleasures and oblivion of the
present: thus 1.7.32, 1.9.13, 1.11.8, 3.29.43, and 4.7.17. In 3.17 Horace again looks ahead to a
private festivity (a celebration of L. Aelius Lamias Genius, somewhat comparable to 3.13, as I
interpret it) from the vantage point of the previous day. If at all marked, then, is cras at 3.13.3 in
actual fact a succinct way of specifying the dramatic time of Horaces song, rather than that of
the sacrifice?
31 Garlands: Varro, Ling. 6.22 (on the Fontinalia); a pig: Mart. 6.47, in explicit payment of a
vow; sheep: Ov. Fast. 3.300; G. Henzen, Acta fratrum Arvalium (Berlin, 1874), 146.
29

138

LLEWELYN MORGAN

or in payment of a negative balance incurred through earlier crimes, though not only
crimes, and itself has symbolic value, standing in as a surrogate for those who offer
it.32 Horace is explicitly offering the fons Bandusiae a gift (donaberis, 3). But the
question would then be, a gift in return for what? For whom or what might an
adolescent goat operate as a repayment or substitute, and what debt might be owed to
the fons Bandusiae to merit such a gift?
One readily available answer might be that the kid is the surrogate of the person
most immediately involved in his sacrifice, Q. Horatius Flaccus, in which case its
symbolic value would potentially be very strong indeed: cut off at the cusp of
maturity, after all, the animal is powerfully evocative of the young Horace himself,
who had taken his leave of the vicinity of the spring to gain an education in Rome and
rise to a life with its fair share of the uenerem et proelia (5) denied to the kid, especially
if Horace is understood in his lyric persona as the Roman Alcaeus, warrior and
symposiast.33 Horaces account of his education at Rome at Serm. 1.6.7182 implies
that his departure from Venusia coincided with the start of the second stage of the
standard elite Roman education, thus placing the future poet at roughly the age of
twelve, a good human analogue for a kid with budding horns.34 A beast full of
youthful promise dies, in other words, in recompense for the success of the boy who
came to write this poem in this book and collection; and the Bandusian spring is
Horaces version of his personal Muses spring, his talent conceived as something
external to himself to which he owes proportionate thanks, in the shape of the
sacrificial kid. The derivation of Bandusia from Greek Pandosia, giver of all,
postulated by Nisbet and Rudd, must also be felt here: return is made to the one who
has given everything.35
I have suggested already that the unsettling quality of this passage is not just a
product of modern sensibilities: Horace was not required to spell out the implications
of the sacrifice for the victim (or for the spring) as explicitly or vividly as he does. Up
to a point this disconcerting emphasis can be satisfactorily explained by spelling out
the logic of such a sacrifice: for the kid to constitute adequate payment for Horaces
success, the future to which he will not attain must encapsulate the life that Horace did
live, the victims lost life a payment for Horaces extraordinary accomplishments. Our
response to the sacrifice is thus not just to mourn the vitality of the haedus (although
we must indeed feel the poignancy of his death, the force of frustra in 6): that lost
vitality also, paradoxically, conveys the rich life with which Bandusia, understood as
32 P.R. Hardie, The Epic Successors of Virgil: A Study in the Dynamics of a Tradition
(Cambridge, 1993), 32, with further references. This way of thinking is particularly important for
the Aeneid, as argued by C. Bandera, Sacrificial levels in Virgils Aeneid, Arethusa 14 (1981),
21739; W.S.M. Nicoll, The sacrifice of Palinurus, CQ 38 (1988), 45972; Ll. Morgan, Assimilation and civil war: Hercules and Cacus (Aen. 8.185267), in H.-P. Stahl, Vergils Aeneid:
Augustan Epic and Political Context (London, 1998), 17597.
33 1.32.512 sets out very clearly the perceived division between Alcaeus political and erotic
poetry, but see also pp. 13940 below.
34 See P. M. Brown, Horace. Satires I (Warminster, 1995), at 1.6.77, with OCD3 s.v. education,
Roman (J.V. Muir).
35 Nisbet and Rudd (n. 15), 1723. A name for a poetic spring translated from Greek also hints
at an important theme in the Odes, Horaces pride in transplanting a Greek poetic form to Italy,
developed most explicitly at 3.30.1314; but cf. 1.20.13 with S. Commager, The Odes of Horace:
A Critical Study (New Haven, 1962), 326, and the suggestive linguistic play at 1.32.34, dic
Latinum, barbite, carmen. The Appian Way, on which the spring in question was positioned, was
of course the link between Rome and Magna Graecia, and thence Greece itself via the port of
Brundisium.

THE ONE AND ONLY FONS BANDUSIAE

139

Horaces ingenium, gifted the poet. With the passage that has been the focus of critical
anxiety, the description of the staining (inficiet) of the spring waters with the blood
of the victim at 68, sacrificial thinking again provides an important key. We cannot
appreciate the richness of Horaces life the generosity of Bandusias gift, in other
words unless the inversion of this gift, the curtailment of life endured by the kid, is
developed with proportionate intensity: and the repayment to the spring of a life for a
life could hardly be represented more compellingly than in the image of a lively
creatures life-force, his blood, seeping back into the springs life-giving waters.
But many readers have still experienced the bleeding of the kid into the spring as
some kind of violation of its purity,36 and although this can be overstated (a spring
adulterated with sacrificial blood is not thereby straightforwardly polluted: sacrificial
blood is at worst an ambivalent substance, as Burkert insists, purifying and polluting
simultaneously),37 Horaces imagery seems carefully designed to render the meeting
of blood and water unsettling. The nuance of the verb inficiet is hard to pin down
with confidence, but the allusion to 3.6.34, infecit aequor sanguine Punico (describing
the actions of the exemplary youth of earlier times), whilst it suggests that the
metaphor might convey the sort of righteous ruthlessness appropriate to crushing
the Carthaginians, nevertheless chimes a little harshly in the quieter, pastoral
environment of the fons Bandusiae. The image is also, most importantly, a contradiction of the Callimachean (and Horatian, cf. 1.26.69 to the Muse, quae fontibus
, the fons
integris / gaudes) ideal of the pure poetic spring (even if debatably
nor integer), and signalled as such
Bandusiae after the sacrifice is neither
by its tension with the opening description of the spring as splendidior uitro: inficiet
(discolour) is set against the purity of splendidior (1).38
In fact this subtle complication of Callimachean programmatic imagery seems as
purposeful as any element of this intricate composition. I return here to the suggestively Alcaic quality of the future denied to the kid, uenerem et proelia. Quintilian
(Inst. 10.1.63) praises Alcaeus for his political poetry, which he considers morally
uplifting in its attacks on tyrants, and in style succinct, lofty (magnificus), precise and
often like an orators (cf. Dion. Hal. Imit. 422), but deplores the same poets
willingness to stoop to frivolous poetry and erotic verse, though more suited to
greater themes (maioribus tamen aptior). Alcaeus is defined as the (regrettably)
(Strab. 13.2.3) and
, light verse alongversatile author of both
side serious, and here, as in Odes 1.32, the essence of the Alcaic persona is the
sympotic poet and participant who is also a man of public affairs, Lesbius ciuis. The
aptness of the Alcaic model to Horace lay above all in the latters own history of
perilous engagement in national affairs, Philippi especially (of which we are about to
be reminded at 3.14.28). What I am proposing is that it matters to Horace to advertise
that his life has been Alcaic, rather than Callimachean, one of proelia (a word with
an inevitable generic charge)39 as well as uenus, and that if indeed Callimachean
36 See the survey of G. Mader, That st(r)ain again: blood, water, and generic allusion in
Horaces Bandusia Ode, AJPh 123 (2002), 519, at 512.
37 W. Burkert, Greek Religion (Cambridge, MA, 1985), 812, with a general statement of the
sacrificial paradox at 81, sacrament and sacrilege merge in every act of sacral killing.
38 Nisbet and Rudd (n. 15), ad loc. As they proceed to comment, this is also one implication of
the figure at 67 by which the redness of the blood implies the clarity of the water, and the
coldness of the water the warmth of the blood: cf. E.A. Schmidt, Schema Horatianum, WS 103
(1990), 5798, at 668. The implied warmth of the blood also contrasts with the delightful chill
that the spring provides against the heat of the Dogstar at 912.
39 See Mader (n. 36), 548 for a comparable reading of 3.13 as a poem in which Horace toys

140

LLEWELYN MORGAN

proprieties are violated, and an edgy hint offered of conflict and violence over and
above that already inherent in the sacrificial ritual, the bloodletting of the haedus
thereby serves to communicate, by inversion, the full richness and complexity of the
Alcaic life to which the Bandusian spring had propelled its protg.40
IV. CONCLUSION: SPRING, POETRY, GENIUS
In the final, vivid evocation of the spring, itself (naturally) an embodiment of
eloquence (1516, unde loquaces / lymphae desiliunt tuae), poetry imitates the spring to
the point of being sonically indistinguishable from it,41 and, more than that, offers in
the very structure of the final stanza a visual reminiscence of the scene described. In a
manner comparable to the poetic landscaping of 3.4.1316, finely elucidated by
West,42 the sense of the text is mirrored in its disposition: the holm oak (14) is set
above the rocks on the line beneath, from which in turn the waters leap down into the
final line, the sinuous movement of sense from line to uneven line iconic of the
tumbling of water over the rocks of the spring:
cauis impositam ilicem
saxis, unde loquaces
lymphae desiliunt tuae.

It would be easy also to draw comparisons between the power of poetry and the
soothing influence of the spring: the frigus amabile bestowed on passing cattle and
flocks (912) recalls the epiphany of the Muse Calliope at the start of Odes 3.4, at
which the poet seems (68) to hear and wander through sacred groves, into which
steal delightful waters and breezes. It is immediately after this that Horace tells of the
magical childhood around such places as Acherontia, Bantium and Ferentum, and
Mt Voltur, on the slopes of which stood Venusia, which marked him out as an acolyte
of the Muses.
Thanks to Odes 3.13 the immortality of the fons Bandusiae does not depend on its
physical location ten miles east of Venosa,43 and that is indeed fortunate: even as early
with a higher poetic register, represented for Mader by, amongst other things, the proelia of 5 and
the motif of bloodstained water. But Mader equates uenus too readily with lyric, regarding the
indications of a higher register of poetry as an epic note (57). But lyric has a much more fluid
generic status than Mader allows. In the person of Alcaeus, especially, not mentioned by Mader,
the polarity of love and warfare is confounded.
40 Lindsay Watson has alerted me in discussion to the parallel between a goat sacrificed to an
inspirational spring and the tale told of Archilochus initiation at testimonia 3 Gerber (SEG
15.517, the Inscription of Mnesiepes), according to which the Muses took a cow from the young
poet-to-be in exchange for a lyre (2235).
41
The ode to Bandusias spring ends with the babbling of water, West (n. 1), 121; cf. Fraenkel
(n. 16), 2034 on the end of the poem (And what an end it is!): Listening to the swift rhythm of
these lines we seem to lose ourselves in the sounds and glitters of an enchanting scenery.
42 West (n. 7), 345.
43 Cf. Joseph Addison, A Letter from Italy, 316, thinking primarily (as Allan Ramsay appreciated, Frischer and Brown [n. 4], 126) of the fons Bandusiae:
Sometimes, misguided by the tuneful throng,
I look for streams immortalizd in song,
that lost in silence and oblivion lie,
(dumb are their fountains and their channels dry)
yet run forever by the Muses skill,
and in the smooth description murmur still.

THE ONE AND ONLY FONS BANDUSIAE

141

as De Chaupys visit the original fountainhead had been obliterated. However, the
geographical facts of the case are far from immaterial, for once we are properly
informed about its true whereabouts we may begin to understand quite why Horace
chose this of all springs to encapsulate his achievements as a lyric poet. Nor is the
profound assimilation of spring and poetry that Horace achieves simply an (unusually
brilliant) instance of something poets typically do. Much rests on the identification
of this Apulian spring and Horaces poetry. For when Horace addresses the fons
Bandusiae, he is contemplating the wonder of his own native genius.
Brasenose College, Oxford

LLEWELYN MORGAN
llewelyn.morgan@bnc.ox.ac.uk

Classical Quarterly 59.1 142146 (2009) Printed in Great Britain

doi:10.1017/S00098388090000111

142
SUSTAINING
DESIRE
MOLLY
PASCO-PRANGER

SUSTAINING DESIRE: CATULLUS 50, GALLUS

AND PROPERTIUS 1.10
In Propertius 1.10, one of many in the monobiblos that address or name a Gallus, the
poet-figure remembers a recent night during which he observed Gallus in an erotic
dalliance with a girl; he thanks his friend for the pleasure and reciprocates with a
poetic offer of erotic aid which is also an offer of aid-through-poetry. The elegy
figures poetry as an appropriate recompense for erotic pleasure, an aid in erotic
pursuit and a medicina for the wounds of love. The relation between poetry and desire
in the poem is further complicated by the suggestion that Propertius seemingly
voyeuristic pleasures are actually textual and not sexual, or are both textual and
sexual, cloaking a description of reading the erotic elegy of Cornelius Gallus.1
Propertius rather creepy night spent2 watching, listening to and revelling in Gallus
and his girls all-night exertions becomes instead a night spent passionately reading
Gallus Amores: the lusus from which Propertius cannot pull himself away (1.10.9)
becomes the play of light verse; the alternae uoces of Gallus and his girl in the heat of
passion (1.10.10) become the alternae uoces of elegiac couplets, or possibly of
amoebean verse.3 Perhaps we are particularly to imagine Propertius reading Gallus
first book, his primus amor (1.10.1); here, in a poem nearly at the centre of his own
first book, Propertius reflects on the sources of his (poetic) passion.4
This critical reconfiguration of Propertius 1.10 as concerned with the erotics of
poetic reception also has brought it into close dialogue with Catullus 50s impassioned
plea to Licinius Calvus. Both Propertius and Catullus have passed sleepless nights (cf.
Cat. 50.10: nec somnus tegeret quiete ocellos; and Prop. 1.10.7, quamuis labentis
premeret mihi somnus ocellos) fuelled by the pleasure of a lusus that is both poetic and
erotic.5 A number of recent studies have explored the homoerotic dynamics of the two
1 A suggestion first made by F. Skutsch, Gallus und Vergil (Leipzig and Berlin, 1906), 1446;
and expanded by A.S. Benjamin, A note on Propertius 1.10: O iucunda quies, CP 60 (1965), 178;
D.O. Ross, Backgrounds to Augustan Poetry (Cambridge, 1975), 834; J.K. King, The two
Galluses of Propertius Monobiblos, Philologus 124 (1980), 21230. See also citations in n. 5,
below. An association of the Gallus of the monobiblos (or at least some of him) with the historical
poet was long closed off by the communis opinio but re-opened in earnest by Ross. Most major
recent studies treat an association of the name Gallus with Cornelius Gallus as part of the
equipment with which a reader approaches Propertius 1. M. Pincus, Propertius Gallus and the
erotics of influence, Arethusa 37 (2004), 16596 offers a summary of the question at 16972.
2 E. Oliensis, The erotics of amicitia: readings in Tibullus, Propertius and Horace, in J.P.
Hallett and M.B. Skinner (edd.), Roman Sexualities (Princeton, 1997), 15171, argues (at 160)
that the testis of line 1 plays with the anatomical meaning of the word and places the speaker
(bizarrely) inside Gallus scrotum; see also Pincus (n. 1) at 1725.
3 On the amoebean reading, J. OHara, The new Gallus and the alternae uoces of Propertius
1.10.10, CQ 39 (1989), 5612, building on J. Fairweather, The Gallus papyrus: a new interpretation, CQ 34 (1984), 16774; contra A. Sharrock, Alternae uoces again, CQ 40 (1990), 5701.
4 On correspondences between this poem and 1.1 (the Cynthia prima): M. Hubbard, Propertius
(New York, 1975), 278; cf. O. Skutsch, The structure of the Propertian Monobiblos, CP 58
(1963), 2389; B. Otis, Propertius single book, HSCP 70 (1965), 144.
5 OHara (n. 3); P.A. Miller, Subjecting Verses: Latin Love Elegy and the Emergence of the Real
(Princeton, 2004), 789; Pincus (n. 1) at 1759. On poetic sleeplessness in both poems: R.F.
Thomas, New Comedy, Callimachus, and Roman poetry, HSCP 83 (1979), 179206 at 2035.

S US TAINING DES IRE

143

poems particularly in their conjunction and have focussed on the figure of the girl /
Cynthia6 in the Propertian poem as a sign of rupture in a Catullan homosocial ideal.
One critic concludes that The inverse dynamics of these two, so similar, poems
correspond to the dramatically different ways in which each text charts the speakers
desire. In the Catullus text, the desire between men is depicted as symmetrical and
dyadic In Propertius, the convergence of poetic production and homoerotically
tinged desire is perverted into a scene of voyeurism by the insertion into the scene of
Gallus puella.7
The bulk of critics are all but silent on the second half of the Propertian poem,
where the poet shifts from reminiscence of his pleasure (iucunda quies iucunda
uoluptas, 14) to the production of a recompensatory gift: accipe commissae munera
laetitiae (12). The same strategy of intertextual reading that has done so much for our
understanding of the first portion of the Propertian poem is nonetheless applicable to
the final section. In particular reading Propertius 1.10.1930 as in dialogue with the
final four lines of Catullus 50 focusses more attention on Cynthia / the puella as
subject rather than symptom in 1.10. The Propertian poem in the end emerges as a
critique of the erotics of poetry played out in Catullus 50 and perhaps also as a
critique of Gallan elegy; most importantly it makes a claim for the self-sustaining
desire-and-poetry produced by Propertius fides to Cynthia.
It must initially be established that the Propertian poem does not abandon its
Catullan intertext after the first ten or twelve lines. The instruction offered by the poet
figure as recompense for his pleasurable reception of Gallus (and his girls) alternis
uocibus, continues the pattern of lexical connection to Catullus 50. Where Catullus
closes with a warning to Calvus not to spurn his poetic offering and his prayers for
continued interaction (nunc audax caue sis precesque nostras, / oramus, caue despuas,
Cat. 50.1819) backed up by a reference to the goddess Nemesis, Propertius poetic
offer of erotic assistance is backed up by reference to Cynthia, who has taught him
what things each person should seek out, and what things he should beware of (quae
cuique petenda / quaeque cauenda forent, 1920; cf. also, tu caue ne tristi cupias pugnare
puellae, 21). Just as Calvus must be careful not to offend Nemesis (laedere hanc caueto,
21), Propertius warns Gallus that an offended girl wont give up her righteous anger
(nec meminit iustas ponere laesa minas).
These fairly mild lexical signs of the connection between Nemesis in Catullus 50
and the figure of Cynthia and the more generalized puella in the Propertian poem are
given more interpretative weight by the broader situational echoes. Nemesis in the
Catullan poem is introduced as an enforcer of fair exchange: she will ensure that
Catullus continuation of the previous days amoebean games is answered in kind.8
6 It does and does not matter whether the girl with Gallus in 1.10 is Cynthia. If the reader is
tracing the narrative of a love affair (and Gallus interference in it), she will remember that in 1.5
Gallus was Propertius rival for Cynthias attentions; 1.13 will find Propertius alone with his love
snatched away (2) and Gallus involved with a girl very like Cynthia, a girl with whom Propertius
has seen Gallus make love. If the reader is thinking instead about literary emulation and rivalry,
Propertius Cynthia is inevitably wrapped up with Gallus amor and his puella. Poem 1.10 itself
recognizes an affinity between the Cynthia of line 19 and the puella of line 5 in that the lessons
learned from the one will apply to the other; they are, after all, both elegiac mistresses,
quintessentially generic women. Cf. King (n. 1) at 214; L. Richardson, Jr., Propertius Elegies IIV
(Norman, OK, 1976), 1801; M. Janan, The Politics of Desire: Propertius IV (Berkeley and Los
Angeles, 2001), at 356; Miller (n. 5) at 84.
7 Pincus (n. 1) at 177. Subtler but similar are Janan (n. 6) and Miller (n. 5) at 67, with further
reference to Oliensis (n. 2) and A. Sharrock, Constructing characters in Propertius, Arethusa 33
(2000), 26384 at 270.
8 D.L. Burgess, Catullus c. 50: The exchange of poetry, AJP 107 (1986), 57686 at 585.

144

MOLLY PASCO-PRANGER

Propertius 1.10s elegiac mistress is likewise positioned as an enforcer of exchange and

particularly of verbal exchange: when Propertius passes on to Gallus what he has
learned under the tutelage of Cynthia, a surprising bulk of the didactic content
concerns the regulation of words: neue superba loqui, neue tacere diu; / neu, si quid
petiit, ingrata fronte negaris, / neu tibi pro uano uerba benigna cadant (Be careful
neither to speak haughtily, nor to keep silent for long; nor, if she asks for something,
should you deny her with a scowl, nor should her kind words fall unheeded by you9)
(1.10.224). Failure to keep up the exchange, to answer with the right sorts of words,
will result in something that sounds very much like the uehemens dea with whom
Catullus threatens Calvus: irritata uenit, quando contemnitur illa, / nec meminit iustas
ponere laesa minas (She comes angered when she is scorned, and, once offended, she
does not remember to put aside just threats) (Prop. 1.10.256). Nemesis is precisely
the embodiment of justified anger10 and the retribution that comes with it, and the
Propertian girl takes up the same position. Her quasi-epiphanic arrival (irritata uenit)
allows her to step all the more easily into the goddess role;11 we may even hear
Nemesis name bubbling under the couplet: contemnitur illa / nec meminit.
The figure of Nemesis, then, is the crux for the intertextual play between the latter
parts of the two poems. Nemesis appearance in Catullus 50.201 has struck many
readers as awkward and out of line with the rest of the poem and indeed there is a
shift in tone in the Catullan poem not unlike that in Propertius 1.10:12 the final lines
are purely monitory with the break strongly marked by nunc at the start of line 17.
More importantly, the introduction of Nemesis as a third figure into the poems
intimate exchange between two men is jarring. In a certain sense the goddess stands in
for the poet: harm me, scorn me, Catullus says, and she will react badly; the repetition
of the verb cauere in lines 1819 and 21 underlines the union between the two. The
interpolation of the goddess allows Catullus to express the prospect of righteous
anger without unduly upsetting the mood of pleasurable intercourse with Calvus.
Still, the shift of tone from all-absorbing passion and longing to warning and veiled
threat, and what seems a sudden triangulation of the two poets mutual exchange
(reddens mutua) reads as a disruption, a break in the hoped-for continuum posited by
the poem. The fact that Nemesis is the primary focus for the intersection of the
didactic portion of Propertius 1.10 with the Catullan poem may in part explain the
frequent modern failure to follow intertextual readings all the way to the end of the
Propertian poem: if critics are unsure what to do with Nemesis, then they may know
even less what to do with Cynthia-as-Nemesis.
9 There is some controversy over the meaning of this clause; is it the uerba benigna of Gallus or
of the girl that should not fall pro uano? S.J. Heyworth, Cynthia: A Companion to the Text of
Propertius (Oxford, 2007), 50 summarizes the positions and comes down on the side of Gallus,
arguing that an interdiction against vain promises on the part of the lover corresponds nicely
with the exhortation in the preceding line to grant the beloveds requests. However, the sense of
continuous exchange of words set up by line 22 is better continued by advising the lover to
respond appropriately to kind words from his girl.
10 Cf. Ov. Met. 3.406, of Nemesis: adsensit precibus Rhamnusia iustis.
11 See Hubbard (n. 4), 26, n. 1, with comparison to Prop. 1.5.32, rogata uenit; cf. also L.A.
Moritz, Well-matched lovers (Propertius 1.5), CP 62 (1967), 1068.
12 A number of critics have read the shift in tone as comical, an intentional over-extension that
tempers the intensity of the Catullan poem: W.C. Scott, Catullus and Calvus (Cat. 50), CP 64
(1969), 16973 at 1712; P. Pucci, Il carme 50 di Catullo, Maia 13 (1961), 24956 at 255; K.
Quinn, Catullus: The Poems (Macmillan, 1970), 239; H.P. Syndikus, Catull: eine Interpretation
(Darmstadt, 1984), 2534; D.F.S. Thomson, Catullus (Toronto, 1997), 326. D. Wray, Catullus and
the Poetics of Roman Manhood (Cambridge, 2001) at 1067 offers a more fruitful integration of
the final lines.

S US TAINING DES IRE

145

If Cynthia / the puella in the second half of Propertius 1.10 plays out the role of
enforcer taken by Nemesis in Catullus 50, that already gives the seemingly dry
didacticism of this portion of the poem more interpretative weight and a closer
relation to the dynamics of exchange in the first half of the poem. In lines 112 Gallus
makes a pleasurable gift to Propertius of the opportunity to observe an exchange of
pleasure and of words between Gallus and his girl; Propertius gives thanks for the
pleasure afforded him by Gallus and promises a fitting gift in return. In 1320
Propertius explains that his gift is not only a recitation of Gallus passions,13 but
something greater: powerful new words, and words learned from his own girl (Cynthia
me docuit). The gift itself follows in lines 2130 and consists of advice on continuing
the proper exchange of pleasure and words between Gallus and his girl. While
Catullus introduces Nemesis at this point, a third figure to enforce the continued
exchange between himself and Calvus, the presence from the start of the girl in the
Propertian poem all but guarantees the continuation of this process of exchange: she
herself will enforce it. A continuation of the one level of exchange will mean a continuation of the other, the repetition of pleasure Propertius prays for in line 4. If we are
taking the erotic pleasure of Propertius 1.10 as a metaphor for, or indeed an analogue
of, poetic pleasure, the girl is positioned as, or perhaps more pointedly recommended
as, the driving force of poetry; the poem concludes with an exhortation to faithful
submission to una puella (2930).14
Of the extant Latin elegy, Propertius poetry, particularly in the Monobiblos,
focusses its elegiac querelae most strongly on the puella in question. Propertius
advertises this unusual focus from the opening of the first poem whose Cynthia prima
is as much a claim to Propertius innovation in the genre as it is a marker of Cynthias
primacy in the poetry. Tibullus had, of course, different mistresses in his two short
books of elegies, Delia and Nemesis,15 and neither is as powerful a force in the poems
as Cynthia is in the Propertian corpus. More importantly for my argument, we have
no indications that Gallus poetry shared Propertius focus on a single beloved: a girlfriend, Lycoris, was clearly present in the poetry, but just as clearly not omnipresent.16
Propertius 1.10, then, which lauds the pleasure Propertius has gained from Gallus
primus amor, but goes on to advise Gallus based on the poets own experience with
13 I here accept Heyworths emendation of line 13 from non solum uestros didici reticere dolores
to non solum uestros didici recitare calores, as printed in the new OCT (2008) and argued in Notes
on Propertius books I and II, CQ 34 (1984), 394405 at 3979.
14 Cf. S. James, Learned Girls and Male Persuasion: Gender and Reading in Roman Love Elegy
(Berkeley and Los Angeles, 2003), which brings to the fore Roman elegys nature as a persuasive
genre directed to a mistress who is constrained by the economic and temporal realities of her
social position as a courtesan; she necessarily resists the impoverished poet in favour of men who
can offer her more concrete resources. Nonetheless, this docta puella has the discernment to be
affected and persuaded from time to time by the offerings of her poet-lover. Her ability (and
indeed her compulsion) to say no with frequency, but also to say a meaningful yes, gives a
self-sustaining impetus to the poetic project of persuasion. Contrast this with, for example,
Millers formulation of Cynthias role in Propertius work: The poetry of the Monobiblos is
inconceivable without Cynthia. She is what allows the work to function and the semiotic game to
be played. Yet she herself never comes into focus; rather, she is like the vanishing point in a
painting that allows the more defined shapes around it to have their form and intercourse with
one another (n. 6 at 66). For Miller, Cynthia is the medium of exchange between men; for James,
she is an agent in the creation of the poetry.
15 Is the name of Tibullus Nemesis perhaps chosen as an acknowledgement of Propertius
recasting of Catullus goddess of exchange as an elegiac mistress?
16 On Lycoris and Gallus: Prop. 2.34.912; Ov. Am. 1.15.2930, Ars 3.5378, Tr. 2.4456; and
Mart. 8.73.510. See Ross characterization (n. 1 at 489) of Gallan poetry.

146

MOLLY PASCO-PRANGER

Cynthia, can be read not just as poetic praise, but as a critique of Gallus, a promise of
something greater than faithful repetition of Gallus mode of poetry. Propertius
claims for his own amor, and his own poetry, an element of self-sustaining exchange.
Throughout Propertius 1.10, as in Catullus 50, words are the primary medium of that
exchange: the proper and pleasurable reception and return of uerba drives this poem
from start to finish. Unlike the Catullan Nemesis, though, the Propertian girl is not
just the enforcer of an exchange of poetic uerba, she is also the stuff of poetry, and
more importantly, a player in the lusus that produces it. Her presence and the alternae
uoces she shares with Gallus are what allow Propertius to enjoy his textual/sexual
pleasure in the early lines of the poem and it is Cynthia who has given Propertius the
words to give back to Gallus. The puellas continued presence means the continuation
of elegiac production and Propertius concluding lines promise more success and
pleasure if Gallus submits to the girl and (therefore) to the continual process of verbal
exchange. Ross tantalizingly suggests that Prop. 1.10 might refer, on one level, to
experiments in amatory elegy Gallus may recently have been writing.17 If so, it claims
(whether chronology allows or not) to be ahead of Gallus in the experiment:
Propertius takes the role of praeceptor amoris to Gallus, and shows him the way to
sustainable erotic elegy.
The Propertian girl thus unifies three roles in the erotic and poetic exchange
sketched out in Catullus 50: she is the goddess / enforcer Nemesis; she is the medium
and material of an exchange of poetic uerba between men, that is, effectively, the
poetry; and most importantly, she herself is a producer of exchanged uerba. The
anguish of the Catullan poem lies precisely in the mediation of the exchange of
pleasure through tablets, verses, metres. When the two men are physically separate the
reality of poetry as medium becomes clearer and it is at this point that anxiety for
continuation and the figure of Nemesis creep in. Propertius recasting of Nemesis and
the union of enforcer, topic and producer of alternae uoces in una puella produce a
more sustainable model of poetic and erotic exchange than the furor of Catullus 50
and perhaps also than Gallan elegy.
The University of Mississippi

17

MOLLY PASCO-PRANGER
mpranger@olemiss.edu
Ross (n. 1), at 83; see also at 102.

Classical Quarterly 59.1 147166 (2009) Printed in Great Britain

doi:10.1017/S00098388090000123

147
ROSALINDE
KEARSLEY
OCTAVIAN
AND
AUGURY

OCTAVIAN AND AUGURY:

THE YEARS 3027 B . C .*
Augustus mentions his transactions with the Senate in late 28early 27 B.C. almost at
the end of his own record of his public career (RG 34.12). The significance of these
sections of his account mean that the recent publication of a new Latin fragment from
Pisidian Antioch is of the greatest importance. This small fragment, containing
lettering from part of three lines, finally reveals that, in 34.1, the participle potitus
restored by Mommsen in 1883 is now to be replaced with the adjective potens:1
1
2

In consulatu sexto et septimo postqua[m b]ella [civil]ia exstinxeram

per consensum universorum [po]tens re[ru]m om[n]ium rem publicam
ex mea potestate in senat[us populi]que R[om]ani [a]rbitrium transtuli.
quo pro merito meo senatu[s consulto Au]gust[us appe]llatus sum et laureis
postes aedium mearum v[estiti] publ[ice coronaq]ue civica super
ianuam meam fixa est [et clu]peus [aureu]s in [c]uria Iulia positus quem mihi senatum pop[ulumq]ue Rom[anu]m dare virtutis clement[iaeque e]t iustitiae et pieta[tis caus]sa testatu[m] est pe[r e]ius clupei
[inscription]em.2

1 In my sixth and seventh consulships, after I had extinguished the civil wars, when through
universal consent I was in possession of power over all public affairs, I transferred the State out
of my power into the authority of the Senate and Roman people.
2 For this service of mine by decree of the Senate I was named Augustus and with laurel
branches the door-posts of my house were wreathed publicly and a civic crown was fixed above
my door, and a golden shield placed in the Curia Iulia, which, it is attested by the inscription of
that shield, the Senate and people of Rome gave to me for the sake of my courage, clemency,
justice and piety.

The discovery has enabled new certainty about the manner in which Augustus chose
to express his position of leadership at a crucial stage of his political career. The
significant position of this episode in a work crafted with the opinion of posterity in
mind (Dio 56.33.1; Suet. Aug. 101.4) suggests that Augustus is describing what he felt
was a landmark in his career. Details that might clarify its political context are
* My thanks are due to E.A. Judge and to the anonymous reader of this article for their
suggestions and comments. It is not intended to imply, however, that they are in agreement with
the views expressed here.
1 The new fragment was first published by P. Botteri, Lintegrazione mommseniana a Res
Gestae Divi Augusti 34,1 potitus rerum omnium e il testo greco, ZPE 144 (2003), 264 and may
also be found together with further discussion in T. Drew-Bear and J. Scheid, La copie des Res
Gestae dAntioche de Pisidie, ZPE 154 (2005), 21760, at 2336. Discussion of RG 34.12 and
the new fragment is also included in J. Scheid, Res Gestae Divi Augusti. Hauts faits du divin
Auguste (Paris, 2007), 8291. The fragment discovered at Pisidian Antioch has confirmed the
earlier proposal to restore potens by R. Kassels (see W.D. Lebek, Res Gestae Divi Augusti 34, 1:
Rudolf Kassels potens rerum omnium und ein neues Fragment des Monumentum Antiochenum,
ZPE 146 [2004], 60). Despite its small size the fragments contribution to the historical analysis
of the text is inestimable, cf. R.T. Ridley, The Emperors Retrospect. Augustus Res Gestae in
Epigraphy, Historiography and Commentary (Leuven, 2003), 2550 and, on RG 34.1 in particular,
13941.
2 The text incorporates the new fragment but is otherwise that of V. Ehrenberg and A.H.M.
Jones, Documents Illustrating the Reigns of Augustus and Tiberius (Oxford, 19552), 28.

148

ROSALINDE KEARSLEY

lacking, however, and it is necessary to turn to other sources in order to investigate the
background to the momentous events to which Augustus refers. Among both contemporary and later sources which are available, the account of this period by Cassius
Dio must be considered of primary importance because it alone provides a chronological framework for events. Despite the considerable distance in time between Dios
own life-time and the events of the Augustan era, the reliability of his chronology for
the Augustan years is strongly indicated by comparison with other sources which also
made use of the acta senatus and the acta diurna.3
In general terms, Dios description of actions taken by Octavian to shore up his
position from 30 down to the beginning of 27 makes it clear that the threat of civil
war was not entirely over in 31, or even in 30 after Antonius and Cleopatra had been
finally disposed of (cf. Dio 52.42.8). Moreover, because of what Dio (51.23.227.3)
reports about the achievements of Marcus Licinius Crassus (cos. 30) and about
subsequent events in Rome, Crassus has been characterized as a rival to Octavian for
pre-eminence in Rome in some modern discussions.4 It is the scale of Crassus military
achievements, the anomalies in how these are celebrated in Rome, and the disappearance of Crassus from the literary sources after the celebration of his triumph,
which have been taken to indicate that he must have been a protagonist in the political
turmoil during Octavians sixth consulship that led to Octavian being given supreme
political power.
The course and nature of Crassus campaigns when proconsul of Macedonia,
provided in the substantial account of Dio (51.23.127.1), surely direct attention
towards Crassus as a potential, and most unwelcome rival to Octavians claim to
military leadership of Rome after 31.5 Moreover, Crassus shifting political allegiances as an adherent of, first, Sextus Pompeius and, then, Marcus Antonius (Dio 51.4.3)
might reasonably be suspected to have contributed both to suspicion of his future
intentions by Octavian, and to a view of him as an alternative leader by contemporaries in Rome, particularly those supporters of Antonius who survived the battle of
Actium and who were once again part of the Roman political scene (cf. Dio 51.2.46,
53.11.14).
In line with such an analysis, the discussion below is based on the following
premises: that the victory of Octavians forces at Actium in 31 did not establish his
right to the leadership of Rome unchallenged; that, during the immediately following
years down to January 27, enmity existed between Octavian and several others who
were determined to lead their lives according to the traditional mores of public life;
finally, that the most difficult and sustained opposition to Octavians determination to
3 M. Reinhold and P.M. Swan, Cassius Dios assessment of Augustus, in K.A. Raaflaub and
M. Toher (edd.), Between Republic and Empire. Interpretations of Augustus and His Principate
(Berkeley, 1990), 1713.
4 See M. Reinhold, From Republic to Principate. An Historical Commentary on Cassius Dios
Roman History Books 4952 (3629 B.C.) (Atlanta, 1988), 162; H.L. Flower, The tradition of
the spolia opima: M. Claudius Marcellus and Augustus, ClAnt 19 (2000), 44 and n. 55 with a
survey of earlier literature. For a recent argument against this view, see J.W. Rich, Augustus and
the spolia opima, Chiron 26 (1996), 85127. A response to his view may be found below.
5 The significance of Crassus thwarted ambition is acknowledged even by K.A. Raaflaub and
L.J. Samons II (Opposition to Augustus, in Raaflaub and Toher [n. 3], 4223) who, in other
respects, minimize the opposition to Octavian between 31 and 28. Perhaps the lengthy excursus
on Crassus operations stems from Dios own interest in the region, but this by no means excludes
the possibility that his sources reflected the interest and admiration Crassus campaigns created
in contemporary Rome. Livys account (Per. 134) has been largely lost.

OCTAVIAN AND AUGURY

149

be sole leader arose as a result of the military achievement of Marcus Licinius

Crassus.

THE YEAR 30
The year following his naval victory over Antonius and Cleopatra was a troubled one
for Octavian. In his account of the year 30, Dio describes how the military veterans
whom Octavian had demobilized and sent back to Italy after Actium began to express
their discontent because he had failed to recognize their contribution to his victory
materially (51.3.12). This discontent gathered strength and led to public demonstrations causing Octavian, says Dio, to be fearful that the soldiers might seek out a
leader and so undermine his recently acquired political control of Rome. The leader
envisaged by Octavian was a potential rival, surely a man of senatorial rank who
would seize on a ready clientela and the opportunity to challenge him for the
leadership. It is a measure of the seriousness of the situation for Octavian that he
surrendered the pursuit of Antonius to his deputies and returned to Italy precipitately
regardless of the fact that it was mid-winter and the sailing season had closed
(51.4.25.1).
Many of all ranks certainly flocked to greet Octavian on arrival in Italy but not all
went spontaneously. Some went because they were summoned and out of fear.
Although the size of Octavians backing in Brundisium on that occasion is unknown,
his arrival must have created the threat of reprisal. Octavian, however, did not attempt
to enter Rome but remained at the port.6
The mutinous veterans were not alone in causing concern to Octavian. Dio also
says there were other urgent matters of business with which Octavian had to deal.
Their political significance is indicated by the fact that neither Maecenas, whom
Octavian had left to protect his interests in Rome, nor Agrippa, whom Octavian had
sent back to Rome from the East in late 31 to ensure that Maecenas authority was
respected (51.3.56), was adequate to handle those urgent matters.7 Some of the
mutinous veterans among the large crowd which had flocked to meet Octavian at the
port received cash payouts and land assignments (cf. RG 16.1) and were, presumably,
satisfied. But, Dio reports (51.5.1), Octavian had to pardon a part of the population
because it had not come to meet him. From this indication that opposition to
Octavian in Rome was maintained by some, we may conclude that the other urgent
business included dealing with it. Marcus Aemilius Lepidus, son of the former
triumvir, who was planning to assassinate Octavian and to revive civil conflict, did not
go unpunished (Vell. Pat. 2.88.13; Dio 54.15.4).8 Clearly, it was to Octavians
advantage to exert himself urgently to prevent the coalescence of disaffected veterans
and any senatorial challengers to his position.
6 Cf. Dio (37.20.36) on how the arrival of victorious commanders at Brundisium created
nervousness among the Romans. As Octavians reception on the occasion of his victorious return
from the East had been planned and announced well in advance (Dio 51.20.13) he may well have
wished to avoid going to Rome before that event.
7 On the contrary, Plutarch (Ant. 73.3) reports that Agrippa kept writing to Octavian that his
presence was urgently needed in Italy.
8 The conspiracy is discussed in P. Sattler, Augustus und der Senat. Untersuchungen zur
rmischen Innerpolitik zwischen 30 und 17 vor Christus (Gttingen, 1960), 2931. The offence of
the equestrian Cornelius Gallus was of a different kind but he too was eventually condemned
(Dio 51.17.1, 53.23.524.1; cf. R. Syme, The Augustan Aristocracy (Oxford, 1986), 32; Raaflaub
and Samons [n. 5], 4235).

150

ROSALINDE KEARSLEY

After what must have been a month of intense negotiation Octavian returned to
Greece (Dio 51.5.1). The nature of the senatorial tributes dating to the latter half of
30 reveals that Octavian prevailed in persuading a majority in the Senate to back him
(Dio 51.19.47). His success was only temporary however. Octavian again suspected
the existence of conspiracies within the higher ranks against him in 29 (Dio 52.42.8).
None of Octavians opponents during the winter of 3029 is named by Dio; among
his supporters, only Maecenas and Agrippa. There is no comment about the behaviour of Octavians co-consul, Marcus Licinius Crassus, at the time of his visit to Italy.
However, Crassus became proconsul of Macedonia later in the same year.9 His war
against the Bastarnae and the slaughter of their king in single-handed combat during
his first provincial campaign won him a triumphal decree by the Senate in the winter
of 3029 (Dio 51.23.224.4).10 This success and its acknowledgement in Rome can
hardly have been welcome news to Octavian.

THE YEAR 29
Octavian was absent from Italy until August during his fifth consulate. Nevertheless,
two events occurred in Rome which Dio says were exceedingly pleasing to Octavian
above all other honours decreed him (51.20.4). These events are of critical importance, therefore, in understanding the character of Octavians relationship to the
Roman State at that time and in the future. The events in question, linked in time by
Dio and dated to the month of January, are the closing of the gates of Janus and the
performance of the augurium salutis.11 Although he was an augur (RG 7.3), Octavians
absence from Rome at that time makes it impossible that he played a role in the actual
performance of that augurium salutis.12 Nevertheless, Suetonius comment (Aug. 31.4)
that it was he who revived the rite must mean that the colleges decision to secure
permission for performance of the augurium salutis from the Senate was at his
instigation.
Even then the augural auspices could be taken only if the augurs were summoned
by a competent official. Octavians consular colleague during the early months of that
year, Sextus Appuleius (Dio 51.20.1), is the most likely candidate both to have
sponsored the request for a decree of the Senate on Octavians behalf and to summon
the augural college for the augurium.13 Appuleius was not only Octavians colleague as
consul ordinarius in 29, he was his nephew.14 Given that he was himself a member of
9 Three suffects followed Crassus as consul (A. Degrassi, Inscriptiones Italiae XIII.1 [Rome,
1937], p. 510), indicating he would have been free to leave Rome by the second half of 30.
10 Crassus campaigns are usually dated 2928 but, although Dio (51.23.12) mentions the
despatch of Crassus to Macedonia under the year 29, he links it with the building of Statilius
Taurus amphitheatre in Rome in 30; cf. Syme (n. 8), 272 and Reinhold (n. 4), 160. For the flexibility of Dios annalistic scheme in recounting foreign wars, see J.W. Rich (ed.), Cassius Dio. The
Augustan Settlement (Roman History 5355.9) (Warminster, 1990), 10.
11 The closing of the gates of Janus is recorded in the Fasti Praenestini (CIL I2.1, p. 231). This
event is proclaimed proudly by Augustus (RG 13, undated) as a mark of his success in achieving
peace and prosperity for Rome, but the augurium salutis is not mentioned.
12 A cippus found below the Arx (ILS 9337) testifies to later ceremonies of the augurium salutis
under Augustus.
13 Most often this was the pontifex maximus (J. Linderski, The augural law, in H. Temporini
and W. Haase (edd.), Aufstieg und Niedergang der rmischen Welt II.16.3 [Berlin, 1986], 2218) but
the political circumstances in 29 made that impossible (cf. RG 10.2).
14 Syme (n. 8), 30.

OCTAVIAN AND AUGURY

151

the augural college,15 it is probable also that Appuleius was the magistrate who took
the auspices and performed the ceremony.16
The last celebration of the augurium salutis prior to 29 was that of the year 63. The
ceremony was held on the successful conclusion of Pompeius extraordinary command in the East against Mithridates.17 When the augurium salutis was revived in 29,
Octavians victory in the East against Cleopatra and the Parthian settlement (cf. Dio
51.17.418.3) would have been similarly represented. The augurium salutis identified
Octavian as the one through whom Jupiter would always act for the well-being and
safety of the State.18 His inseparable link with the safety of the Roman State at an
official level is documented by the inclusion, in 29, of his name in the public hymns
with those of the gods (Dio 51.20.2). Among these Augustus later singles out for
special mention the hymn of the Salii (RG 10.1).19 In the same year Octavian cleansed
the temple of Capitoline Jupiter of all earlier dedications and replaced them with his
own (Dio 51.22.34; Suet. Aug. 30.2). By this means, he further identified himself with
the chief god of the State, and hence with the fortune of the State itself.20 Thereafter,
he laid the laurels from his fasces in the Capitoline temple after each victorious war
(RG 4.1; cf. Dio 54.25.4). The first words of an Augustan edict relayed by Suetonius
(Aug. 28.2) displays his self-identification with the divinely-granted security of the
State: ita mihi salvam ac sospitem rem publicam sistere in sua sede liceat (May I be
permitted to maintain the State safe and sound in its proper position).21 Recognition
of this important role of Octavian may be found in the subsequent traditions of the
inhabitants of Augustan Rome and in Augustan literature (cf. Virg. Aen. 1.28696).22
It is surely also reflected in Asia Minor where Octavian was at the time of the
augurium salutis (Dio 51.20.751.21.1). There, dedications linked his name with that
of Hygeia (Salus).23
15

T.R.S. Broughton, The Magistrates of the Roman Republic (Atlanta, 1952), 2, 532.
On the procedure of the augurium salutis, see Linderski (n. 13), 22556.
17 Dio 37.24.125.2. Like Octavian, Pompeius had not yet returned to Italy. The similarity in
the political circumstances leading to the auguria salutis of 63 and 29 should not pass unnoticed.
However, an important difference between them is that the earlier of the two was irregular. When
repeated it was again accompanied by unlucky omens and Cicero later reported Appius Claudius
as saying the irregularity meant civil war was at hand (De div. 1.47.105). No such cloud hung over
the ceremony of 29.
18 Cf. Linderski (n. 13), 2226, 2291; J. Linderski, Roman Questions. Selected Papers (Stuttgart,
1995), 490. The role of Augustus in this respect is also depicted on the gemma Augustea where he
is shown in the guise of Jupiter sitting beside Roma as he is crowned with the corona civica.
Significantly, he holds the augurs lituus rather than the thunderbolt of Jupiter (see P. Zanker,
trans. A. Shapiro, The Power of Images in the Age of Augustus [Ann Arbor, 1988], 2301 with 234,
fig. 182).
19 Ovid, Fasti 3.36192 reveals the significance of this for claiming divine endorsement of his
leadership of Rome.
20 Given the tributes to Octavian of the year 36 and the connection then made between his
victory and Capitoline Jupiter (Dio 49.15.12), the augurium salutis must have presented the
people of Rome with confirmation of the existing relationship rather than a novel concept. The
view of E. Fantham (Rewriting and rereading the Fasti: Augustus, Ovid and recent Classical
scholarship, Antichthon 29 [1995], 53) that Augustus took no personal initiative to enhance the
cult of Jupiter Optimus Maximus or his own association with Jupiter seems unlikely in the light
of these events.
21 Not surprisingly there is no indication of the date of this edict in Suetonius.
22 Throughout his life Octavian/Augustus health was of importance to Romes inhabitants (cf.
Suet. Aug. 57.1).
23 C. Habicht, Inschriften von Pergamon 8, 3 (Berlin, 1969), 1645; J. Reynolds, Further information on imperial cult at Aphrodisias, StudClas 24 (1986), 109 = SEG 30 (1980), 1269.
16

152

ROSALINDE KEARSLEY

Octavians revival of the augurium salutis should not be considered only as part of
his interest in restoring discontinued rituals of Rome. A specific political purpose was
served by this particular revival.24 The augurium salutis was used by Octavian as a
means towards ensuring sole and enduring control in Rome. Dio says (51.20.5) that
the closing of the gates of Janus and the augurium salutis took place despite the fact
that Roman forces were still campaigning in Germany and elsewhere. The Senates
action in sanctioning the ceremony meant that it endorsed his campaigns as the only
significant ones for the well-being of the State. Its decree that the gates of Janus
should be closed and prayers for future safety could be offered similarly had the effect
of diminishing the significance of the activities of other members of the senatorial
class who were still campaigning.25
Of the Roman commanders known to have been active outside Italy in 3029, it is
Licinius Crassus whose military exploits must have demanded a response from
Octavian.26 As early as his first season of campaigning as governor of Macedonia,
Crassus had distinguished himself among Roman commanders in an extremely rare
manner. He had captured the armour of the enemy king in single combat (Dio
51.24.4).27 For Octavian, this achievement must have represented the emergence of a
dismaying threat to his own desire for continued supremacy. He could not afford to
allow Crassus to surpass him in the strictly measurable sense that dedication of the
spolia opima signified.28 If that occurred there was a strong possibility that his own
successes against Cleopatra and the Parthians would be overshadowed (cf. Dio
51.19.120.3).
An extra dimension to the situation was created by the fact that Caesar had
signalled his intention to campaign in the Thraco-Macedonian region against the
Getae on his way to attack the Parthians (Appian, Ill.Wars 3.13). Since Caesar was
prevented from doing this by his premature death, his heir may well have considered
that task his prerogative.29 Moreover, Crassus had apparently intruded upon a client
relationship of the Caesarian family with Roles, a king of the Getae (cf. Dio
51.24.67, 51.26.1). Thus, family honour may have been at stake, giving Crassus
achievement the character of a personal as well as a political rivalry in the eyes of
Octavian. Whether or not this was the case, the fact remains that full recognition of
Crassus achievements would place him in a position from which he could challenge
Octavians right to the sole leadership of Rome. Hence the crucial importance
to Octavian of the augurium salutis of January 29. It secured for himself alone the
right to decide if the State was at war by the sacred role of interpres of Jupiter. No
doubt it was this outcome that made performance of the ceremony so pleasing to him.
24 Cf. Octavians revival of the ritual of the fetial priesthood in 32 (R.A. Kearsley, Octavian in
the year 32 B.C.: the S.C. de Aphrodisiensibus and the genera militiae, RhM 142 [1999], 5660).
25 Dio 51.20.5 may reflect a justificatory explanation of Octavian, perhaps contained in his
Memoirs.
26 See J.J. Wilkes, The Danubian and Balkan provinces, in A.K. Bowman et al. (edd.), The
Cambridge Ancient History2 X (Cambridge, 1996), 54853. Dio preserves contradictory traditions about the nature of Crassus operations (cf. 51.20.5 with 51.23.227.3). The contemporary
campaigns of Carrinas are mentioned but not described in Dio 51 21.6.
27 Dio 51.25.12 reveals that Deldo was dead and the triumph announced before the winter of
3029 had set in. Crassus news was probably known to Octavian already by the winter of 3029
since he was also in the East (Dio 51.21.1).
28 On the spolia opima, see Rich (n. 4), 106; Flower (n. 4), 501.
29 Cf. Reinhold (n. 4), 66. Octavians announcement of a campaign in Britain in 27 (Dio
53.22.5) reflects his propensity to emulate his adoptive father. Despite diplomacy replacing
military action, he claims success there in RG 32.1, as does Horace on his behalf (Od. 3.5.24).

OCTAVIAN AND AUGURY

153

Octavians return to Rome was followed closely by remarkable triple triumphal

celebrations in mid-August.30 The triumphal procession of the third day was
especially noteworthy for the manner in which the usual order was reversed to
Octavians advantage and the detriment of his magisterial colleagues (Dio 51.21.9).31
Two triumphs, one for the naval victory at Actium and one for the conquest of Egypt,
had been decreed in the previous year (Dio 51.19.15). To these a third was added and
celebrated as the first of the three days. Appian refers to it as an Illyrian triumph
(Ill.Wars 5.28), and so it is usually described in modern discussions, as Appian attributes this triumph to a long-delayed celebration of Octavians own campaigns in
Illyricum between 35 and 33. Dio describes the day as celebrating victories over the
Pannonians and Dalmatians, the Iapydes and their neighbours, and some Celts and
Galatians (Dio 51.21.57). According to him, the day was actually a composite
occasion on which Octavian caught up under his own name Carrinas victories to the
north of Italy (Dio 51.21.6), Crassus achievements over the Dacians and the Moesians
(Dio 51.25.2), and his own earlier campaigns in Illyricum against the Iapydes (Dio
51.21.5). Although Crassus name is not included in Dios description of the day, the
claims made for Octavian by Appian bear a far closer resemblance to Crassus
campaigns (cf. Dio 51.27.1) than Octavians own achievements in Illyricum.32 Moreover, in the procession of that day, as also on the other two days, spoils from Egypt
were carried (Dio 51.21.7). Thus Octavian used the first day of his triple triumph to
display the results of others victories and to assert his primacy by displaying Egyptian
spoils at the same time.
Octavians celebration of the victories of others as his own would not have been
unusual during the period of the second triumvirate. Then, commanders triumphed
only by permission of Octavian and Antonius.33 Triumviral prerogatives were not
needed by Octavian to justify the continuation of this practice after Actium however.
Nor did he need to seek some re-definition of precedence for consular imperium over
that of a proconsul.34 His precedence was based upon his membership of the Iulian
family and the honours inherited by him as the adopted son of Iulius Caesar.35
30

Degrassi (n. 9), p. 570.

Cf. Reinhold (n. 4), 158.
32 Reinhold argues (ibid. 74, 157) that Appians text is heavily dependent on the Memoirs of
Augustus. A variety of sources indicates that captives from Crassus first campaign were
exhibited by Octavian on that occasion and at the consecration of the shrine of Iulius Caesar in
the same year (Dio 51.22.23). Virg. Aen. 8.728, for example, mentions the presence of hitherto
unconquered Scythians at the triumph who, when combined with information in Dio 51.22.6, RG
31.2 and Livy, Per. 134, should probably be recognized as Dacians defeated by Crassus. Certainly
Dio (51.22.8) observes that some of these prisoners were captured after the battle of Actium.
They cannot have been from Octavians own campaigns therefore: cf. A. Mcsy, Der vertuschte
Dakerkrieg des M. Licinius Crassus, Historia 15 (1966), 512. Zanker (n. 18), 689 identifies a
visual depiction of the first day of the triple triumph in the frieze of the temple of Apollo
Sosianus where bound, northern, barbarians are seated to either side of a trophy with enemy
helmet and armour (70, fig. 55). The trophy depicted in the lower register of the gemma Augustea
may allude to the same occasion although Zanker connects it with Tiberius later campaigns
(ibid. 2312, fig. 182). K. Galinsky, Augustan Culture (Princeton, 1996) 107 and fig. 44 points to
the helmeted trophy depicted on the rear of the cuirass on the Prima Porta statue of Augustus but
makes no attempt to explain its presence.
33 Dio 49.42.3 makes it clear that negotiation and bargaining were part of this process.
34 This possibility was suggested by R. Syme, The Roman Revolution (Oxford, 1939), 309; cf.
also Linderski (n. 18), 614. However, a general in the pre-Augustan period conducted a war under
his own auspices and, provided his auspices were valid (cf. ibid. 61617), that man was the
commander-in-chief. Hence, in the years 3028 Crassus was in fact commander-in-chief (so
Flower [n. 4], 52; Ridley [n. 1], 86).
35 Dio 48.46.1 states that Caesar did not avail himself of all the honours voted him by the
31

154

ROSALINDE KEARSLEY

In 45 a decree of the Senate in favour of Caesar, his sons and his grandsons had
defined the relative importance of Romes commanders and the nature of their
awards for victorious campaigns. The Senate granted Caesar the right to celebrations
for victory even if he had not been on the campaign or had any hand at all in the
achievements (Dio 43.44.6). Octavian claimed this right and, during the second half
of 29, spent extravagantly with the aim of deepening public veneration of his adoptive
father (Dio 51.22.12). Thus, in the political climate of the period there was nothing
inherently unusual in Octavians outranking of Carrinas and Crassus in the procession on the first day of his triple triumph.
The eventual celebration of a triumphal procession by Crassus himself makes it
unlikely that acknowledgement of the imperatorial acclamation was withheld from
him as Dio states (51.25.3).36 The contradiction created by Dios statement, therefore,
can be understood only in light of the senatorial endorsement of the Caesarian family
in the year 45. For at that time the Senate decreed that Caesar was to be Imperator
once for all. He was given the right to bear the title permanently as a proper name, as
were his sons and grandsons in turn (Dio 43.44.23; Suet. Caes. 76.1). This award
involved a new way of using the designation Imperator, but it did not thereby replace
the old. The imperatorial acclamation of the soldiers for their commander was not
abolished by the Senate. The permanent and the acclamatory usages of Imperator
were to co-exist (Dio 43.44.45). While there is disagreement among modern scholars
as to whether or not Caesar himself availed himself of his new nomenclature,37 no
disagreement exists with respect to Octavians use of Imperator as a praenomen.38
Dio highlights the year 29 when dealing with the use of the praenominal title by
Octavian (52.41.34), but Octavian had actually not delayed so long in taking up the
inheritance. Octavian availed himself of the hereditary right to this award soon after
Caesars death. It is attested almost immediately after the treachery of Salvidienus
Rufus in 40.39 Octavian demanded harsh punishment of Salvidienus in the Senate,
asserting his offence extended beyond betrayal of personal friendship to treason
against the entire Roman people. The basis of his claim must have been his possession
of the name Imperator as heir of Caesar. The Senate accepted the gravity of the
situation as portrayed by Octavian and passed a senatus consultum ultimum (Dio
48.33.3). Hence, when Dios narrative for the year 29 attributes the denial of Crassus
right to dedicate the enemy kings armour to Octavians superior imperium (Dio
51.24.4), he is actually recording the re-assertion of the right to the praenomen
Imperator by Octavian following an effective precedent of a decade earlier.
The prosecution of a prolonged and major war during his second campaign in the
Macedonian region (Dio 51.25.327.3) delayed Crassus return to Rome and the
celebration of his achievements in the field.40 Octavian attempted during this time to
Senate in 45. This in no way affects Octavians right to choose to use them later on; cf. Reinhold
(n. 4), 2312.
36 Degrassi (n. 9), p. 87. Cf. L. Schumacher, Die imperatorischen Akklamationen der
Triumvirn und die auspicia des Augustus, Historia 34 (1985), 210; Ridley (n. 1), 86.
37 Cf. R. Syme, Imperator Caesar: a study in nomenclature, Historia 7 (1958), 1767;
M. Gelzer (trans. P. Needham), Caesar: Politician and Statesman (Oxford, 1969), 307, n. 2.
38 Cf. Syme (n. 37), 179.
39 H.A. Grueber, Coins of the Roman Republic in the British Museum (London, 1910) vol. II,
pp. 41011, nos 1014; cf. Syme (n. 34), 113, n. 1. For Salvidienus revolt: Appian, BC, 5.6; Dio
48.33.23.
40 Crassus return was probably late in 28 (E. Badian, Crisis theories and the beginning of
the Principate, in G. Wirth [ed.], RomanitasChristianitas: Untersuchungen zur Geschichte und

OCTAVIAN AND AUGURY

155

consolidate the Senates support for himself. A review of the senatorial roll was begun
by Octavian and Agrippa as censors during 29. The record of discontent in the
senatorial class lacks detail but is not entirely extinguished. Octavian caused some
senators to depart and he enrolled other men, even adlecting some of them to consular rank (Dio 52.42.14). According to an authorization of the Senate and People
given in the previous year,41 he created more patricians to fill the ranks of important
traditional priesthoods (RG 8.1; cf. Dio 52.42.5). He sought to win senators over with
the promise that he would burn incriminating letters found among Antonius
possessions (Dio 52.42.8).42 None of these measures can have stemmed the senatorial
opposition to Octavian because he also curtailed the senators right to travel outside
Italy without permission, except to Sicily and Gallia Narbonensis (Dio 52.42.67). It
is not hard to imagine what prompted this action. In the year 32 a large proportion of
the Senate had caught Octavian unaware and left Italy to join Antonius in the East. In
29 Octavian may have anticipated, or at least feared, a repetition of such an event, this
time in Crassus favour.
Octavian not only worked to marshal the senatorial order to his support during 29,
he also invested much money and energy in developing political support in other
sections of Roman society. He spent a considerable amount of the proceeds from
Egypt on the citizens resident in Rome (RG 15.1). Military veterans benefited, too, by
cash and by land (RG 15.3). In the same year he returned to the municipia and colonies of Italy their contributions of 35,000 pounds of aurum coronarium (RG 21.3).
Ultimately, however, Octavians claim to superiority had been most specifically
expressed to all of those in Rome by including Crassus victories of 3029 in the first
day of his triple triumphal celebration while Crassus was still on campaign in
Macedonia. After that any attempt by Crassus to dedicate the armour of Deldo as
spolia opima on his return would amount to a specific confrontation with Octavian (cf.
Dio 51.24.4).
Nevertheless, Octavian did prepare for that eventuality. With specific reference to
the issue of the spolia opima, Livy records that Octavian claimed he, rather than
Crassus, held the auspices for the campaign when King Deldo was killed in single
combat (4.20.67). Octavians claim did not have any legal basis since, as proconsul
of Macedonia, Crassus imperium and auspices would have been independent of
Octavian for the period he spent within his province.43 Certainly there is no indication
that Octavian was alleging that a vitium had been uncovered in the departure auspices
of Crassus nor indeed that there had been auspical irregularity in the field at the time
of his campaigns. Therefore, since Octavian was not asserting that the magisterial
auspices of Crassus prevented him from dedicating the spolia opima, his claim must
have been based on the auspices he held as a result of the augurium salutis.44 The
Literatur der rmischen Kaiserzeit [Berlin, 1982], 25, n. 19, 26). The date is not indicated by Dio.
However, he refers to three major expeditions undertaken by Crassus after that of 30 against
Deldo, namely those against the Maedi and Serdi and the provision of help to Roles
(51.25.426.1), the siege of Ciris (51.26.36), and the attack on the Artacii (51.27.1). He also says
specifically (51.27.2) that a considerable time had passed.
41 His authorization was the Lex Saenia: Reinhold (n. 4), 212.
42 This promise was not completely fulfilled and men were condemned on the basis of evidence
found there later on (Dio 52.42.8). Sattler (n. 8), 314 clarifies the nature of the difficulties
Octavian faced in seeking to establish support for himself in the Senate.
43 See n. 34 above; cf. Linderski (n. 13), 221718; Schumacher (n. 36), 213.
44 Cf. Linderski (n. 18), 61011. The terms auspicia and auguria were sometimes used interchangeably: see H.D. Jocelyn, Urbs augurio augusto condita: Ennius ap. Cic. de Div. I. 107 (=
Ann. 7796 V2), PCPS 17 (1971), 489; cf. Linderski (n. 13), 2292.

156

ROSALINDE KEARSLEY

armour of Deldo was won during activities unsanctified by Jupiter in Octavians view.
Crassus trophy could not be recognized as spolia opima, or so Octavian must have
publicly argued in 29.45
J.W. Rich has also argued against deeming that Crassus held an imperium which
was inferior to Octavians and missed out on the imperatorial salutation for his
campaigns.46 With respect to the spolia opima, however, Rich is prepared to allow for
the possibility that Crassus deferred to Octavian, or even that he refrained from
claiming the right to dedicate the spolia opima unprompted.47 Apart from the fact that
such abnegation is uncharacteristic of the Roman nobility, the weakness of Richs
argument lies in the fact that he fails to identify in this context the significance of the
earlier endorsement of Octavian through the augurium salutis. Nevertheless, in
acknowledging the propensity of Crassus military success to eclipse and to distract
from Octavians own victories, he does recognize the strong grounds Octavian must
have used to argue against Crassus celebration of the ritual of the spolia opima. An
objection by Octavian on the basis of the augurium salutis could not have been
considered trivial.48 Octavian had been acknowledged as the interpres of Jupiter by it.
It is because of what was at stake for each man that a political clash must have
occurred.49 Crassus achievement extended far beyond the usual level of military
valour among the Roman aristocracy. The feat of the spolia opima had been achieved
by only three Romans before 29, the first of whom was Romulus.50 The spolia opima
and the related topic of personal bravery, then, must have been a volatile issue in
Rome even going into 28. It must have marked out Crassus as Octavians most
dangerous opponent.51 It was an intricate situation for Octavian because the historical
tradition might be interpreted in Crassus favour.52 Hence, it would be understandable
if Octavians actions during the following year while Crassus return was awaited were
designed to distinguish himself in the life of the State and to leave no room for a
challenge to his leadership by Crassus.
THE YEAR 28
Octavian was consul for the sixth time in 28. Significantly, Agrippa was his only
colleague and the pair remained in office throughout the year. The year-long appointment of Agrippa as his consular colleague, although suffect consuls had become
45 Cf. E. Badian, Livy and Augustus, in W. Schuller (ed.), Livius. Aspekte seines Werkes
(Konstanz, 1993), 1516 against the common view that Livys notice of Augustus opinion
(4.20.7) was an insertion written after 27; contra Rich (n. 4), 118.
46 Ibid. 937.
47 Ibid., 107.
48 Cf. ibid. 1006, 1267.
49 Ridley (n. 1), 15960 believes that in Crassus Octavian faced the most serious challenge to
his military standing during 28.
50 For the nature of the tradition and those who dedicated the spolia opima before 29, see
Flower (n. 4), 3448.
51 Badian (n. 40), 26 seriously underrated the political acuteness of Octavian when he argued
that no one would have thought of the significance of Crassus winning the spolia opima during
29. In a later discussion, Badian (n. 45), 15, did recognize the extreme political difficulty any such
dedication by Crassus would cause Octavian. Cf. also, Flower (n. 4), 50, 53.
52 Ibid. 52; Rich (n. 4), 93. If Dio is correct in reporting the Senate had awarded Caesar the
right to dedicate spolia opima in the temple of Jupiter Feretrius in 45 (44.4.3), Octavians proprietary sense with respect to Caesar and his privileges may once again have been aroused by
Crassus. Dios information is sometimes regarded as unreliable but cf. Flower (n. 4), 489. Livys
personal confusion (4.20.78) has preserved the political ambiguities which existed.

OCTAVIAN AND AUGURY

157

customary, is evidence that Octavian was finding the political environment in Rome
particularly difficult that year (cf. Dio 51.3.57, 52.1.2).53 Nevertheless, Dio describes
the general character of Octavians actions from the beginning of 28 as being deeply
respectful of ancestral custom (cf. RG 8.5). He illustrates this by describing what he
sees as distinctive aspects of his leadership at that time (53.1.13), namely Octavians
collegial attitude towards Agrippa his co-consul by sharing the fasces, the swearing of
the customary oath at the conclusion of his consular year and the revival of the
principatus of the Senate.
Octavians designation as princeps senatus appears to represent the achievement of
a senatorial consensus. The latest attested principatus for Republican Rome was that
of Cicero in 43 when he was designated as princeps senatus and recognized as princeps
Romani nominis.54 The latter title, equivalent in meaning to princeps Romanae
civitatis, was an acknowledgement of the standing which had underlain the designation princeps senatus since 209.55 Thus, as princeps senatus, Octavian was ostensibly
acknowledged as leader of Rome by the senators. The position gave expression in the
civic sphere to the auctoritas he had gained through the augurium salutis on the basis
of his military achievements in the previous year and was of great political significance. However, the circumstances of Octavians appointment were not necessarily
indicative of a voluntary affirmation of him. While Dio does not relate by whom or in
what circumstances Octavian was designated princeps senatus, it is most likely that his
appointment was made jointly by the censors, that is by Octavian himself and
Agrippa.56 Since appointments were commonly associated with a lectio,57 the appointment probably occurred prior to the lustrum Octavian performed on completion of
the censorial revision (cf. RG 8.2).
The other actions of Octavian discussed by Dio (53.12) under the year 28 involved:58 the distributions of funds (to impoverished senators to enable them to hold
public office; to the plebs a four-fold distribution of grain, and to the aerarium funds
as necessary); measures of restoration and reconciliation (the cancellation of many
debts owed to the aerarium before the battle of Actium; the destruction of records of
debt to the State, and the rescinding by edict of all unjust and illegal laws he had
sponsored when a triumvir to take effect from the end of his sixth consulship);59
attention to cult (the completion and dedication of the Palatine temple of Apollo;
provision for the maintenance of Roman temples, and the expulsion of Egyptian rites
53 Agrippas first consulship was in 37 and this was also a crucial year for Octavian (Dio
48.2.4). Rich (n. 4), 110, in disregard of Dios account, argues that Augustus was totally secure in
his leadership during the years between the victory over Antonius and the beginning of 27 (cf.
Ridley [n. 1], 15960).
54 F.X. Ryan, Rank and Participation in the Republican Senate (Stuttgart, 1998), 2023.
55 Ibid. 2301.
56 Ciceros appointment was unconventional in that he was designated by the Senate (ibid.
203). Nevertheless, the political context of April 43 reveals how that appointment too served a
political purpose.
57 Ibid. 2323.
58 The events of this year appear to be narrated thematically rather than in sequential order, cf.
Dios words
in 53.2.4. Arguably this is a reflection of the haphazard survival of
the history of this crucial period, something also indicated by the extreme brevity of Dios
account of 28 and the lack of information about it in the Res Gestae.
59 An aureus was issued in 28 with the legend LEGES ET IVRA P R RESTITVIT, see
J.W. Rich and J.H.C. Williams, Leges et Ivra P.R. Restitvit: a new aureus of Octavian and the
settlement of 2827 BC, The Numismatic Chronicle 159 (1999), 17688, at 180. According to its
legend the coin is most likely celebrating this edict.

158

ROSALINDE KEARSLEY

from within the pomerium). In addition, Octavian provided at his own expense the
first celebration of the festival decreed in honour of the victory at Actium. It took an
elaborate form and included chariot races for young and old among the nobility,
gymnastic contests for which a special stadium was built, and also gladiatorial
combats between prisoners of war.
By these actions Octavian portrayed himself once again, to the entire social order
in Rome, as patron and benefactor as well as leader. Moreover, the Actian festival also
revived the triumphal atmosphere of the preceding year. It and the dedication of the
temple of Apollo offered Octavian the opportunity to remind those in Rome of his
victory over Antonius and the Egyptian queen. His action with respect to Egyptian
cult within the city provided a tangible illustration of his mastery over Egypt in 30.
The highly cultivated and well-rounded image of leadership which must have resulted
from the total of Octavians actions of 28 was of vital tactical importance to him in
the face of Crassus successful campaigns and winning of the spolia opima.60 Dio
records that indeed these measures did earn Octavian popularity and praise (53.2.6).
There are other significant events that took place during the year which Dio does
not record. The fact that the Senate issued a decree authorizing Octavian to restore
82 temples of Rome which were decaying and needed repair during 28, his sixth
consulship, is known only from the Res Gestae (20.4).61 But Ovid does refer to
Octavian/Augustus as builder and restorer of temples and by doing so on the Kalends
of February (Fasti 2.5564) he most likely provides that decrees date.62
Given his renewed eminence with respect to cult during 28, Octavian may well have
considered that his new responsibility for the religious life of the city should include
the performance of a lustratio on its behalf. Such a ceremony would have provided the
perfect context for him to adumbrate the potential danger arising from Crassus
impending return to Rome.63
The end of the census was traditionally marked by a purificatory lustratio and
Augustus himself records that, as censor, he did carry out that ceremony (RG 8.2).64
60 Rich and Williams ([n. 59], 183) link the issue of the aureus to the settlement of 27, a view
based on their prior assumption that the transfer of power was a process extending over 2827
not as a single act in January 27; cf. nn. 97, 103 below for a different view. (The question of
whether P. R. in the legend should be expanded by the genitive or the dative [Scheid (n. 1), 86] has
no impact on the relevance of the coin in either context.)
61 Ridley (n. 1), 1823 expresses puzzlement at the formulation of this paragraph and surmises
that Augustus meant the date only to apply to the Senates authorization. If this is the case, it
highlights the significance of the decree in Augustus eyes. It is also noteworthy that the
rebuilding resulting from this decree is distinguished by Augustus, both by inclusion of this detail
and by his arrangement of the text, from the many other individual restorations and constructions he carried out (cf. RG 1920.3; 21.1). Dio (53.2.4) may constitute a less specific allusion to
the decree and to a range of other related measures taken by Octavian.
62 No other reason for Ovids commemoration of Augustus on this day is known, cf.
G. Herbert-Brown, Ovid and the Fasti. An Historical Study (Oxford, 1994), 43.
63 For the threatening circumstances in which a lustratio might be considered appropriate, cf.
Lucans account (1.584609) of an extraordinary lustratio in 49 B.C. Horace (Od. 4.15), describing
how the threat of attack from the Dacians and Getae was averted, conceivably reflects an
atmosphere of fear in 28 which had been engendered by Octavian.
64 The so-called Altar of Domitius Ahenobarbus may well represent this scene. Identification
of the date and circumstances of the altars creation has so far eluded scholars (cf. R.M. Ogilvie,
Lustrum condere, JRS 51 [1961], 37; E.S. Gruen, Culture and Identity in Republican Rome
[Ithaca, 1992], 14550, pls 23). Arguably, however, the representation of censorial activity
including a sacrifice to Mars, and marine friezes honouring Neptune, combined on the one
monument, fit well with Octavians celebrations, both in 28, of the naval victory at Actium by the

OCTAVIAN AND AUGURY

159

The Amburbium, a purificatory festival instituted by Numa, was regularly celebrated

on the Kalends of February (Macrob. Sat. 1.13.3).65 Octavian may also have performed a lustratio on that occasion, especially since it is the same date as the Senates
edict authorizing him to restore the citys temples. As a pontiff (cf. RG 7.3) he was
eligible to officiate at the Amburbium. Moreover, contemporary sources repeatedly
draw a parallel between Octavian and Numa (Virg. Aen. 6.81011; Livy 1.19.3).66 In
fact, since the exact date of the censorial lustrum is unknown, it is possible that
ceremony was on the Kalends of February and it was extended to include the circular
procession so that the entire city was purified. Certainly, Augustus (RG 8.2) describes
his censorial lustrum by the phrase lustrum facere, which had a wider application than
the technical one, lustrum condere.67
The connection of Octavian with another ancestral ritual early in 28 is suggested
by Ovids reference to Augustus strengthening of Romes wall (facit hic tua magna
tuendo moenia) on the Nones of February (Fasti 2.1334), a ritual known as either
augustum augurium or inauguratio.68 This augural ceremony was associated with
several of the revered figures of Romes regal past. It had been performed by Romulus
at the founding of Rome (Livy 1.7.14).69 Numas inauguration as king had provided
the basis for his new foundation of the city on law and ancestral custom (Livy
1.18.619.1).70 Thus an inauguration might equally relate to the well-being and
prosperity of the city in peace as to the definition of its boundaries and consolidation
of its defensive position.71 The depiction of Octavian/Augustus as wall-strengthener
by Ovid denotes the designation of him also as a Founder through the same ritual.72
Once again, then, pious action offered the opportunity for Octavians name to be
linked with that of Numa, to Octavians own political advantage. In 28 the performance of an augustum augurium would have been a striking way to associate himself
with Romes well-being for the purposes of short-term political gain. Over later years,
Augustus decisions would certainly increase the citys prosperity by organization,73
dedication of the temple of Apollo and new Actian festival (cf. the inscription on the victory
monument at Actium: R.A. Gurval, Actium and Augustus: the Politics and Emotions of Civil War
[Ann Arbor, 1995], 667, n. 116 where Neptune and Mars are mentioned together) and the
ceremony at the conclusion of the censorship of himself and Agrippa (RG 8.2).
65 G. Wissowa, Amburbium, RE 1 (1894), 181617.
66 Cf. Herbert-Brown (n. 62), 49; R.J. Littlewood, Imperii pignora certa: the role of Numa in
Ovids Fasti, in G. Herbert-Brown (ed.), Ovids Fasti (Oxford, 2002), 17580.
67 For detailed discussion of the difference between lustrum condere and lustrum facere, see
Ogilvie (n. 64), 31.
68 The phrase augustum augurium is used by Ennius apud Varro, Rust. 3.12. Livy (5.52.2) uses
inauguratio and its cognates for the same ritual: Jocelyn (n. 44), 489.
69 See Jocelyn (n. 44), 4950.
70 Littlewood (n. 66), 180.
71 Littlewood (n. 66), 179.
72 Octavians action recalls Numa rather than Romulus since it is unlikely from Ovids
description that Octavian extended the pomerium. Servius Tullius had also earned the title
Founder without extending the pomerium by his census and distribution of citizens into classes
and centuries (Livy 1.42.35; cf. Littlewood [n. 66], 179, n. 13). Whether or not, at some time
later, Augustus did extend Romes pomerium remains a debatable point. Tac. Ann. 12.23, Dio
55.6.6, and SHA Aurelian 21.11 all record he did so, but it is not mentioned in the Res Gestae or
the Lex de imperio Vespasiani (CIL VI.930); cf. Ridley (n. 1), 82.
73 Augustus is known to have divided the city into a larger number of regions than Romulus.
The existence of fourteen regions is first attested in 7 B.C. (Dio 55.8.7) and this may well be the
date of the final Augustan organization. It is unlikely to be the date of the earliest action of
Octavian/Augustus in this respect, however, as the total of fourteen was probably only arrived at
as part of a process of systematic expansion (cf. J. Bleicken, vici magister, RE VIIIA [1958],
2481).

160

ROSALINDE KEARSLEY

by construction and restoration (Suet. Aug. 2830), and by the attainment of peace
over many years in the future. Thus the analogy between the accounts of Numas
actions and the strengthening of Romes civic and social structure by Octavian/
Augustus continues to be notable even after 28. Octavian/Augustus public image was
transformed over the years from a factional aggressor to the sacred protector of the
State.74 It is surely no coincidence, therefore, that it was also on the Nones of
February (Fasti 2.11927) that Augustus accepted the title pater patriae in 2 B.C., or
that it is this occasion that he represents as the crowning achievement of his public life
(RG 35.1).75
No literary source records that an augustum augurium was performed by Octavian.
Neither is there a reference to one by Augustus in the Res Gestae. It is only the
evidence of a denarius from the East that indicates its occurrence and date. The
connection Livy draws (4.20.7) between the firmly dated restoration of the temples
and Octavian as Founder also points to the same conjunction of ceremony and its
date.
The denarius in question was long ago identified by Gag as bearing a representation on the reverse of Octavian performing an inauguratio.76 He argued that Octavian,
shown with veiled head and ploughing a furrow with a pair of oxen, is depicted in the
role of Founder, although not in the literal sense like Romulus but rather as Founder
in an allusive sense, as a Second Founder.77 A date of between 29 and 27 was
assigned to the coin at that time.78 But now, when the combination of the inauguratio
type on the reverse with the depiction of Octavian as Apollo on the obverse is
considered, the date of the issue may be estimated even more precisely. Since Octavian
dedicated the temple of Apollo adjoining his house on the Palatine in 28 (Dio 53.1.3),
it is surely to that year, above all, that the assimilation of Octavian with Apollo is to be
expected.79
In all it is probable that during 28 Octavian acted repeatedly to demonstrate he was
the one who guarded Romes relationship with its gods and, through the rituals of
lustratio and augustum augurium in particular, to convey the message that peace in
Rome was under threat. His emulation of Numa in the various ways discussed above
would have contributed immeasurably to the impact of this strategic behaviour. As
well as enabling Octavian to gain in stature as the perpetuator of ancestral custom (cf.
RG 8.5; Dio 53.1.1), it allowed him to align himself specifically with the other great
Founder of Rome besides Romulus. In doing this he achieved a tactical balance with
Crassus who stood in the tradition of Romulus after winning the spolia opima. And,
finally, his message of impending threat to Rome meant that those in Rome viewed
him as the one most fit to preserve the citys well-being and prosperity.
Towards the end of 28 after the dedication of the temple of Apollo on 9 October,80
74

Littlewood (n. 66), 192.

Suet. Aug. 58.
76 J. Gag, Les sacerdoces dAuguste et ses rformes religieuses, MEFR 48 (1931), 934.
77 The image of a priestly figure ploughing a furrow refers to the ritual appropriated by Rome
from the Etruscans. It defined the pomerium of a new settlement (Varro, Ling. 5.143; Aulus
Gellius 14.1.3); cf. Ovid, Fasti 4.81819 for the ploughing action of Romulus. For a list of other
conditores after Numa who also only founded figuratively, see Littlewood (n. 66), 179, n. 13.
78 H. Mattingly, Coins of the Roman Empire in the British Museum (London, 192364) vol. I,
p.104, no. 638, pls 15, 17.
79 Cf. Zanker (n. 18), 829; O. Hekster and J. Rich, Octavian and the thunderbolt: the temple
of Apollo Palatinus and Roman traditions of temple building, CQ 56 (2006), 1625.
80 CIL I2.1, p. 331. The Actian festival was associated with this.
75

OCTAVIAN AND AUGURY

161

Octavian was in a strong position to consolidate further political support for himself
against Crassus by the use of invective.81 Octavian could denigrate Crassus by
drawing attention to his political vacillation during the civil wars. Perhaps even his
consulate of 30 and his appointment to Macedonia might be cited to show Crassus
expediency in striking a bargain after Octavian had defeated Antonius.82 Crassus was
also vulnerable to innuendo over his familys impiety. The disastrous result of his
grandfathers Parthian campaign was attributed to the disregard of inauspicious
omens (Plut. Cras. 18). Survival of the story as a topos of wrong behaviour towards
the gods was probably given new life by Octavian at this time.83
Furthermore, the tremendous success of Crassus operations in the second
campaigning season could also have been turned to his disadvantage by Octavian.
The regions of Macedonia, Illyricum and Thrace were the location for armies and
battles during the civil wars on more than one occasion.84 They formed the bridge
between Italy and the wealth of the East and were also of strategic importance for
Italys security against foreign enemies. In the decades after Actium the Romans
feared Italy might be invaded from that area (Virg. Georgics 2.4978; Hor. Od.
3.8.1718; Hor. Sat. 2.503). In 28 Crassus was in control there and had legions at his
command. Moreover, former Antonian supporters had been settled in Macedonia
after Actium (Dio 51.4.6). These men could be represented as a potential source of
additional manpower for Crassus should he wish to turn on Rome.85
Nearby native tribes would readily come to a Romans mind as potential military
allies for Crassus also. The significance of diplomatic links with tribal kingdoms is
revealed by the frequency with which client relationships are referred to and native
embassies are recorded as visiting Rome or entering into treaties of alliance.86
Relationships of patronage were developed by Roman leaders among these tribes
including apparently by the Licinii.87 Crassus is recorded as having friendly relations
with King Roles of the Getae (Dio 51.24.6, 51.26.1) and with the Odrysae (Dio
51.25.45) during the period in question.88 His influence extended to the cities of
Greece. He was the patron of Thespiae, an important city of southern Boeotia, which,
81 For the probability that Crassus return was late in 28, see n. 40 above. Octavian waged a war
of words with great effect against Antonius in 32, cf. Kearsley (n. 24), 54.
82 Cf. E. Groag, Licinius, RE 25 (1926), 2712. The record in Dio 52.42.4 that C. Cluvius and
C. Furnius were adlected among the ex-consuls in 29 because their year of office had been taken
by others reveals the nature of such accommodation and negotiation.
83 Hor. Od. 3.5.58; Dion. Hal. 2.6.4; cf. S.P. Mattern-Parkes, The defeat of Crassus and the
just war, CW 96 (2003), 393.
84 Wilkes (n. 26), 5459.
85 After the division of provinces in 27, Macedonia was left with little by way of a military
garrison even though its border was repeatedly challenged by outsiders (Syme [n. 8], 2745). The
removal of the legionary force reflects Augustus later suspicion of any Roman commander with
a large force so close to Italy.
86 Cf. Dio 51.22.8, 51.24.7; Front. Strat. 1.10.4; Plut. Ant. 63.4; Suet. Aug. 63.2. Compare the
discussion of G.W. Bowersock, Augustus and the Greek World (Oxford, 1965), 1525 for the
period of Pompey and the second triumvirate.
87 A large clientela of the Licinii existed in Illyrian Istria in the time of Nero, one which Tacitus
described as being of long standing (Hist. 2.73).
88 The trial of Primus for making war on the Odrysae less than a decade later (Dio 54.3.2)
illustrates how relationships with this tribal group still aroused strong feelings in senatorial
circles. The depth of political division on the matter is demonstrated by the plot that was formed
against Augustus after the trial (Dio 54.3.4) as well as by the level of his own celebration when the
plot was extinguished (Dio 54.3.8).

162

ROSALINDE KEARSLEY

together with Athens, honoured him and acknowledged his goodwill.89 These cities
also recorded the imperatorial acclamation Crassus had won during his campaigns in
Macedonia.90
Remnants of Octavians verbal attacks on Crassus may still be discernible in the
speech to the Senate which Dio attributes to him in January 27.91 For example, in
53.7.1 Dio portrays Octavian as claiming conquests in Moesia which actually belonged
to Crassus. Elsewhere (53.7.34), Dio gives Octavian the words of a self-denying
leader who stressed the contrast between himself and others whose passion for sole
supremacy meant they would even use violence to win it. Direct evidence for the other
side of the debate is harder to discern in the sources. However, Octavians repudiation
of his most unpopular acts during the triumviral years (Dio 53.2.5), and the other acts
of civilitas (Dio 53.2.13, 53.2.67), suggest he was forced into a defensive posture
during 28 by the reality of Crassus opposition and support for it in Rome.
The scale of publicly expressed antagonism between these two leading senators
would undoubtedly have led to fear and uncertainty among all Romans. Civil war had
been the result of such situations more than once in recent history.92 It was this which
created the climate in which a lustratio would have been widely supported by the
population of Rome.93 In addition, Octavians augural role as mediator between
Jupiter and Rome continued to provide him with the authority to interpret the gods
will about any threats he perceived against himself or the city.94 The level of concern
among senators and people alike would have favoured an attempt by Octavian to seize
the initiative without even waiting for Crassus return.95 When writing about the
performance of the augurium salutis, Cicero had argued that a good augur cooperated
with Jupiter not only in protecting but also in rescuing the State in emergencies (De
leg. 3.43). An argument from Octavian that an emergency situation was threatening
the well-being of the state would, therefore, carry the weight of the auspices.96
Such supreme power had been decreed for Octavian by an SCU as recently as 40
when, following his denunciation of Salvidienus, the Senate committed the care of the
city to the triumvirs (Dio 48.33.13). Given Octavians designation as princeps senatus
at the beginning of 28, obtaining a majority vote in the Senate for an emergency
89 Thespiae: AE 1928.44; Athens: IG2 III.4118. Thespiae had good relations with Rome over a
long period and had a community of Roman businessmen. Cf. M. Kajava, Cornelia and Taurus
at Thespiae, ZPE 79 (1989), 148.
90 Their testimony is inestimable as they record a fact which disappeared from the Roman
historical tradition after Augustus was established as Romes leader.
91 Cf. the comments of Reinhold (n. 4), 84 on the speeches of Antonius and Octavian included
by Dio in Book 50.
92 The analogous situation between Pompeius and Caesar in 49 (Dio 41.3.34) would also still
have been fresh enough in the public mind.
93 The number of citizens Augustus claims to have registered at the lustrum in 28 (RG 8.2) is
believed to be impossibly large if it involved only male citizens (P.A. Brunt and J.M. Moore
(edd.), Res Gestae Divi Augusti. The Achievements of the Divine Augustus (Oxford, 1967), 51. The
likely inclusion of women and children suggests a wider purpose for the occasion than the traditional registration of the citizen roll; cf. n. 67 above.
94 Linderski (n. 13), 22956 underlines the fact that, in contrast to the auspicium of magistrates
which was valid for one day only, the augurium was permanent since it expressed Jupiters will for
a person, place or ceremony, and not merely the timing of an action.
95 Cf. the manner in which Cicero undermined Antonius in 43: Phil. (Cic. Orations, trans.
W.C.A. Ker, Loeb Classical Library vol. 15 [Cambridge, MA, 1926]); cf. Dio 45.1847.5,
especially 45.446.2.
96 Cf. Linderski (n. 13), 22535 for the inclusion of an invocatio uti avertantur mala in the
augurium salutis.

OCTAVIAN AND AUGURY

163

power directing him to ensure the city came to no harm would not have been difficult
to achieve (cf. Dio 53.2.7).97 There is no record in Dio of such an official endorsement
of Octavian. However a cistophoros minted in Ephesus and dating to 28 appears to
reflect the successful outcome after the award of supreme power to Octavian. It bears,
on the obverse, the laureate head of Octavian together with the legend LIBERTATIS
P.R. VINDEX (champion/defender of the liberty of the Roman people) and, on the
reverse, an image of Pax. Wording analogous to this legend, rem publicam ... libertatem vindicavi, is found in RG 1.1 where the context is also civil war. From this
Augustus own usage and from traditional usage, the intended allusion of the coin is
made clear.98 It provides contemporary evidence for factional division in Rome in 28,
division which was believed to be so threatening to Rome that Octavian was called
upon to rescue the State. It also records Octavians eventual political victory over
Crassus.99
The preservation of the memory of this battle of wills was undesirable from
Augustus point of view once he had prevailed. Crassus military achievements were
quickly minimized. He did triumph some six months later for victories ex Thraecia et
Geteis but there was no mention of the victories over the Dacians, although this was
Crassus major success.100 The celebration is likely to have been presented publicly as a
magnanimous gesture of clemency by Octavian.101 Octavians clemency, if such it was,
did not extend to the dedication of the spolia opima (Dio 51.24.4; cf. Livy 4.20.67).102

CONCLUSIONS
Written many years later, the formulation of RG 34.1 refers to the Senates ultimate
deferral to Octavian as leader in the vaguest possible terms and offers no indication of
the events of 28 which preceded it.103 Interpretation of the text by modern scholars
97 For the view that an emergency power was conceded to Octavian in 28 and was released
back to the Senate before the end of the year, see E.A. Judge, Res Publica Restituta. A modern
illusion?, in J.A.S. Evans (ed.), Polis and Imperium: Studies in Honour of Edward Togo Salmon
(Toronto, 1974), 279311, 293 with 309, n. 16.
98 The coin was most probably minted between mid October and the end of 28 (cf. Badian [n.
40] above). Details of this coin are given in Rich and Williams (n. 59), 171. Cf. also W. Eder,
Augustus and the power of tradition: the Augustan Principate as binding link between Republic
and Empire, in Raaflaub and Toher (n. 3), 8997.
99 Cf. Judge (n. 97), 294. E. Ramage, The Nature and Purpose of Augustus Res Gestae
(Stuttgart, 1987), 68 with n. 149 and Rich and Williams (n. 59), 1836 detect a reference to
Octavians victory at Actium in this legend. However, the language used is not in accord with the
consistent characterization of that conflict as a foreign war on coins and inscriptions (cf. ibid.
171; Gurval [n. 64], 667, n. 116).
100 Wilkes (n. 26), 550 suggests that this was ignored because it was offensive to Octavian; cf.
RG 30.2 and 31.2 which appear to be Augustan appropriations of Crassus victories.
101 Crassus triumph: Degrassi (n. 9), p. 87. Dio 56.38.35 describes the political use of
clemency by Octavian after Actium. The theme of clemency was traditionally associated with the
civic crown (Judge [n. 97], 2923). Hence, clemency was appropriate for the manner in which the
conflict between Octavian and Crassus was resolved. The gift of clemency by Octavian after
prevailing in no way prejudices the likelihood of enmity between the two men. Although the lack
of traces of a public debate in the sources is attributed by Rich (n. 4), 108 to the absence of any
confrontation between Octavian and Crassus, it is more probable that the official sources no
longer acknowledged it had taken place (cf. Ridley [n. 1], 93).
102 A Marcus Licinius Crassus is consul in 14 B.C. (Dio 54.24.1) but, whether or not this is a
son (doubted by Syme [n. 8], 276; accepted as an adoptive son by Flower [n. 4], 50), it does not
illuminate the subsequent career of the cos. 30.
103 The interpretation that a specific threat does underlie RG 34.1, however, is supported by

164

ROSALINDE KEARSLEY

was also complicated for many years by the gap in the Latin text where there once
stood a single but crucial word. Only very recently has Augustus own portrayal of the
political situation in 2827 been clarified by the fragment discovered at Pisidian
Antioch (n. 1 above). It is now clear that Augustus described himself as potens rerum
omnium. By selection of the word potens, Augustus intended to make clear the
military and political nature of his supreme power.104 This is in accord with the
character of the entire account of his achievements which emphasizes the military
aspects of his power above any other.105 Even though he did not name his opposition
of 28, it was obviously a matter of pride to Augustus that the context of his civic
awards of January 27 be appreciated. Despite the fact that the words per consensum
universorum failed to make explicit, or perhaps deliberately concealed,106 the actual
source of his supreme power they do indicate Augustus view of the importance of the
period in his public career.107
The cistophoric coin minted in Octavians sixth consulship expands and confirms
such conclusions drawn from the Res Gestae. It cannot have been a minor political
matter which resulted in the naming of Octavian as vindex libertatis populi Romani
that year. On the contrary, this epithet, like potens rerum omnium in the Res Gestae,
relates to a political crisis of great moment and must reflect a genuine struggle for the
hearts and minds of the Roman people by Octavian against strong opposition.108 It
has been suggested that the legend on the cistophoros reflects the passing of a decree
of the Senate honouring Octavian.109 If this was indeed the case, it is likely that it was
only from this point on that Octavian felt sufficiently secure in his leadership of the
State to surrender his supreme power to the Senate.
The amount of information preserved in the sources about the honouring of
Octavian as Romes saviour on 13 January in 27 forms a strong contrast with the lack
of recorded detail about the year 28 (RG 34.12; Dio 53.16.4). Nevertheless, a link is
established by the way the rewards the Senate gave Octavian/Augustus refer back to
the threatening atmosphere of civil war. The wreathing of the door-posts of his house
with laurels, and the civic crown placed above his door, symbolized a triumph over
enemies and fellow citizens rescued.110 Pompeius had felt that winding laurel around
reference to Tiberius recapitulation of the events of the years from 31 to 27 in his funeral speech
for Augustus (Dio 56.34.441.9). Tiberius refers to a period after Actium which would have been
one of factional strife had not the Senate insisted Octavian accept their choice of him as leader
again (56.39.56). The theme of Octavians recusatio, prominent in this part of the speech as it is
elsewhere in Dio (cf. 53.5.14), should not be allowed to distract from the political discord which
is the topic of discussion (cf. Dio 53.11.14).
104 See the discussion of the meaning and usage of potens in Scheid (n. 1), 826, at 85.
105 Ridley (n. 1), 92.
106 Cf. Ridley (n. 1), 109 on Augustus concealment of the nature of his censorial power of 28
in RG 8.2.
107 For the use of comparable emphasis to highlight other significant events, cf. RG 5.2, 9.2,
10.2, 25.2 (and Kearsley [n. 24]), 35.1. Scheid (n. 1), 86, by contrast, attributes a political significance to such expressions of support.
108 Ridley (n. 1), 221 argues that the total control of Octavian was an ongoing position
following his victory at Actium. However against this, in addition to the difficult political situations which Octavian faced (discussed above), there is the nature of the rewards for Octavians
surrender of power (see RG 34.2 with n. 110 below).
109 Rich and Williams (n. 59), 1867. It is far less likely that an honorific decree was issued in
the middle of gradually letting go aspects of the leadership (cf. ibid. 191) than at the successful
completion of a particular action which had been previously endorsed by the Senate.
110 These accolades from the Senate (RG 34.2) have a character directly linked to victory in war
and therefore cannot be explained simply as honorific tributes; see Judge (n. 97), 2904.

OCTAVIAN AND AUGURY

165

his fasces was an inappropriate way of expressing victory over fellow citizens (Dio
41.52.1). Caesar, too, hesitated to appear to rejoice publicly after his victory over the
Pompeians (Dio 42.18.12). Not so Octavian. Moreover, after the Senates affirmation
of him as leader on 13 January, he sought approval for a preferential rate of pay for
his future Praetorian Guard (Dio 53.11.5).
This is not how events were presented by Augustus in his Res Gestae (345)
however. He re-wrote history when he described the award of his new name Augustus
in isolation from its full historical context. Despite the acknowledged connection of
his name with augury,111 the Res Gestae is silent about the auguria of 2928. Augustus
deliberately shifted focus in the Res Gestae from 2928, difficult years, to the Senates
endorsement of him early in 27.112 Yet the absence of any reference to the augurium
salutis or the augustum augurium is anomalous since much of his auctoritas during the
years 3027 stemmed from his consummate use of augury. The augurium salutis had
played a central role in defining his military activity as, alone, important for Romes
prosperity and in identifying him as Romes mediator with Jupiter in guarding the
peace of Rome. The augustum augurium had permitted the pairing of himself with
Numa and his designation as a Second Founder of Rome.113
In RG 34.2, Augustus writes as if there was a causal connection between the award
of the name and his transfer of power back to the Senate. The description of the
award of his new name at that point was intended to associate it with a period of
security and political unity in Rome rather than one of insecurity and division. He
does this also by mentioning his new name before listing the honours of the laurels,
civic crown and the shield.114 But the true sequence of events was different.
Dio (53.16.7) describes the debate about Octavians new name as taking place only
after the Senate had recognized his unique value to the State by the award of honours
and the division of powers. By revealing the passage of time before the award of the
name Augustus to Octavian, Dio shows that it was not a reward for the return of
the extraordinary power to the Senate.115 It was the military honours which were the
rewards for Octavians surrender of the extraordinary power. Statements like that of
Horace (Od. 3.6.116) which attribute the establishment of peace in Rome to religious
piety are not a true reflection of the basis of Octavians supremacy even though they
are significant in demonstrating how quickly an Augustan interpretation of the
period took root. Peace did not come immediately after Actium; it was actually won
after a struggle and sustained conflict between Octavian and some of his senatorial
peers. Peace came only after a period of extreme tension which eventually forced the
Senate and people of Rome to make the choice of Octavian over Crassus at the end of
28.
111 This connection was in the public arena already during the lifetime of Octavian (Ovid, Fasti
1.60712). Its etymological validity is unassailable (Linderski [n. 13], 22901).
112 Abbreviation and arrangement of information, or even total silence, is used repeatedly by
Augustus in the Res Gestae to bypass people or events from his past which he intended should
disappear from the historical record; cf. the fragmented treatment of the triumviral years (RG 2,
7, 25 and also Ridley [n. 1], 23441).
113 It is not surprising that the name Augustus, which recalled Numa with his peace-loving
strengths rather than war-like Romulus, was felt by Octavians supporters to be the more appropriate for him as Founder (cf. Dio 53.16.68.; Suet. Aug. 7.2; Florus 2.34.66).
114 Cf. the comments of Judge (n. 97), 289 on the non-chronological arrangement of the
honours in RG 34.2. Scheid (n. 1), 901 points out that the linking of the gift of the clupeus aureus
with the other gifts in RG 34.2 may represent a conflation also.
115 CIL I2.1 p. 231 (Fasti Praenestini) provides epigraphic evidence to the same effect.

166

ROSALINDE KEARSLEY

As a result of this new standing, and using his augural authority, Augustus could
henceforth even re-interpret calamities. Contrary to precedent (cf. Dio 50.8.3) and
inconsistently with future interpretations (cf. Dio 53.33.5, 54.1.1), the Tibers flooding
on the night when he became Augustus was said to point to a period of new success
for him as Romes leader (Dio 53.20.1).116 And Augustus continued to anchor his
leadership in augury. The frequent inclusion of augural symbols on visual monuments from the period of his leadership of Rome clearly reveals how the augural
doctrine remained central to Augustan ideology.117
Macquarie University

ROSALINDE KEARSLEY
r.kearsley@humanities.mq.edu.au

116 Linderski (n. 18), 61819 cites this as an example of the manner in which Augustus
conquered state religion.
117 Development of the priesthoods and ritual of the cults in the 265 vici of Rome (cf. Zanker
[n. 18], 12934) was probably intended by Octavian/Augustus as a complement to his inauguration.

Classical Quarterly 59.1 167186 (2009) Printed in Great Britain

doi:10.1017/S00098388090000135

167
MONUMENTS
AND MEMOR
Y
GEOFFREY
S. SUMI

MONUMENTS AND MEMORY: THE AEDES

CASTORIS IN THE FORMATION OF AUGUSTAN
IDEOLOGY
I. INTRODUCTION
When Augustus came to power he made every effort to demonstrate his new regimes
continuity with the past, even claiming to have handed power in 28 and 27 B.C. back to
the Senate and people of Rome (Mon. Anc. 34.1). He could not escape the reality,
however, that his new monarchical form of government was incompatible with the
political ideals of the Republic. At the same time, Augustus was attempting to reunite
a society that in the recent past had been riven by civil conflict. It should be no
surprise, then, that the new ideology that evolved around the figure of the princeps
attempted to retain the memory of the old Republic while at the same time promoting
and securing the power of a single authority through which Rome could flourish.1
The new regimes relationship to the recent past was complicated, too, inasmuch as
Augustus power was forged in the cauldron of the late Republic, and he was the
ultimate beneficiary of the political upheaval of his youth. Augustus new ideology
had to recall the Republic without lingering over its tumultuous last generation; it had
to restore and renew.2
Augustus boast that he found Rome a city of brick and left it a city of marble as
well as the long list in the Res Gestae (Mon. Anc. 1921.2) of monuments that he
either built or restored declare that the new topography of the city was an important
component of this new ideology. One monument can serve as a case study: the Aedes
Castoris in the Forum Romanum. This temple, which dated back to the very
foundation of the Republic and had more recently been the locus of intense and often
violent political confrontations in the era before Augustus came to power, was
rededicated in A.D. 6 in the names of Augustus successor Tiberius and his deceased
brother Drusus. This rededication, I shall argue, was part of a larger process through
which the memories associated with the temple helped shape the ideology of the
Principate. One critical juncture for the formation of this new ideology was the
succession, since it was only when power was promised or bequeathed to a successor
that the reality of monarchy was revealed. In order to understand how this process
worked, we shall focus on three memories evoked by the Aedes Castoris: first, the
temples political function in the late Republic, especially as a location for public
meetings (contiones) and legislative assemblies (comitia), made it resonant of popular
sovereignty; second, the temples foundation myth, which told of the Dioscuris
appearance at the battle of Lake Regillus, and further the traditional mythology of
the Dioscuri, were adapted for the succession; third, the temple also served as the
destination point for the transuectio equitum, a Republican ceremony of great
importance to the equestrian order and revived by Augustus in the Principate in part
as a showcase for his successors. Romans maintained a dialogue with the past through
1 A. Gowing, Empire and Memory: The Representation of the Roman Republic in Imperial
Culture (Cambridge, 2005), 201.
2 Cf. K. Galinsky, Augustan Culture: An Interpretive Introduction (Princeton, 1996), 288.

168

GEOFFREY S. SUMI

memories evoked by this and other monuments in the city. Under the Principate,
however, the main interlocutor in this dialogue was the princeps himself, and his
ideology was the filter through which many of these memories were communicated to
the Roman people.3

II. MONUMENTS AND MEMORY

Before we can discuss in detail the place of the Temple of Castor in Augustan ideology, it is necessary to establish the theoretical underpinnings on which this discussion
will rest. For Romans much of their history was contained in the monuments that
dotted the landscape of their city, and these monuments, I would argue, acted as a
kind of mnemonic device that allowed Romans to remember some of the great events
of their past.4 The process of remembering is complex and dynamic. At one end is the
memorial, the physical reminder, which often stands in the place of person or persons,
or even events, now remote in place or time.5 However, the Roman habit of constantly
restoring and refurbishing temples, not to mention the more mundane task of
maintenance and preservation, had the effect of keeping the founders memory alive6
and by the same token the memory of any event or events with which the monument
was associated. At the other end of the process of remembering is the viewer who
brings to the memorial his own complex of ideas and attitudes that inform the
memory produced. In the case of ancient Rome, we are usually hard pressed to do
better than merely guess at the precise content of this complex of ideas and attitudes,
but we are certainly witness to examples of when it is exploited.
One such example is the penchant of Roman orators for using monuments as visual
aids or props in their speeches (Cic. Scaur. 468, Cat. 4.18, De or. 2.2667; cf. Quint.
Inst. 5.10.41). In his first speech against Catiline (1.33),7 which was delivered at a
Senate meeting, Cicero used the setting of this speech, the Temple of Jupiter Stator, to
cue his audience to think about the early history of Rome. At the very end of the
speech, Cicero mentions the founding of the temple by Romulus and lingers over the
significance of the epithet Stator, Stayer. Ciceros mention of Romulus here was
meant to draw the audiences attention to the foundation myth of the temple and
allow him to compare himself with Romulus.8 Cicero thus became another stayer at
another crisis in Roman history.
Ciceros rhetorical ploy could be effective only if he and his audience shared the
3 Cf. J. Rea, Legendary Rome. Myth, Monuments, and Memory on the Palatine and Capitoline
(London, 2007), 8.
4 On monuments and memory, see Varro, Ling. 6.49, with the discussion in M. Jaeger, Livys
Written Rome (Ann Arbor, 1997), 1518. Cicero can speak of the power of suggestion that exists
in places (vis admonitionis inest in locis, Cic. Fin. 5.2). The study of memory in Rome has
exploded in recent years; in addition to Jaeger (above), see, in particular, C. Edwards, Writing
Rome (Cambridge, 1996); U. Walter, Memoria und Res Publica (Frankfurt am Main, 2004);
Gowing (n. 1); Rea (n. 3); and D.H.J. Larmour and D. Spencer, Introduction Roma, recepta: a
topography of the imagination, in D.H.J. Larmour and D. Spencer (edd.), The Sites of Rome:
Time, Space, Memory (Oxford, 2007), 160.
5 Cf. Walter (n. 4), 133, who emphasizes the vastness of time that often separates monument
and memory.
6 Walter (n. 4), 136.
7 On the invocation of Jupiter in this section of the speech, see A.R. Dyck (ed.), Cicero,
Catilinarians (Cambridge, 2008), 1223.
8 As A. Vasaly has persuasively argued (Representations: Images of the World in Ciceronian
Oratory [Berkeley, 1993], 4159).

MONUMENTS AND MEMORY

169

same memory associated with the temple and its founding. Ciceros audience on the
occasion of his speech was clearly cued by the very words of the orator cued to think
about Romulus and the founding of the Temple of Jupiter Stator. In other words, one
memory contained within this monument was evoked by Cicero for a particular
purpose on a particular occasion. In other cases, oral tradition and historiography,
inscriptions and coins, might help to evoke a memory contained in a monument.9
Many monuments, however, had more than one foundation myth, or contained several
memories, some of which were in direct conflict with one another and therefore must
have complicated the process of remembering. The Lacus Curtius, for example, recalls
the death of one Curtius in the war between Sabines and Romans but was he a
Roman hero or Sabine enemy?10 Tacitus appears to be exploiting the conflicting
memories evoked by the Lacus Curtius to underscore the ambiguity of the regime and
death of Servius Galba, whom he describes pointedly as perishing at this very
location.11 Cicero in his speech against Catiline, and Tacitus in his narrative of
Galbas death both show how places can evoke memories. While Cicero directs his
audience explicitly to the memory that he wants them to recall, what Tacitus does is
much more subtle and suggestive: he merely points to the location of Galbas death
and requires his readers to recall the necessary memories. Both methods, as we shall
see, were at work in Augustan ideology.
III. THE AEDES CASTORIS AND POPULAR POLITICS IN THE
LATE REPUBLIC
The Temple of Castor was more than a cult centre and became increasingly important
in the political life of the late Republic, as a meeting place for the Roman Senate (Cic.
Verr. 2.1.129), the destination point in the annual parade of knights (transuectio
equitum) (see further below), and especially as a site of contentious public meetings
(contiones) and legislative assemblies (comitia) orchestrated by politicians or the
urban crowd.12 In order to understand how the Temple of Castor came to be
associated with popular politics, we must first consider the place of the temple in the
changing topography of the Forum Romanum. Then, we can move on to show how
control of the temple came to be equated with control of the popular assemblies that
met there and hence control of popular sovereignty. We will conclude with a discussion of the temples political function under Augustus.
The temple became the site of public meetings and legislative assemblies only after
the Forum changed its orientation from the early to the late Republic.13 The focal
9

Walter (n. 4), 136.

Livy 1.12.10, 13.5; 7.6.16; on the Lacus Curtius, see the discussion in D. Spencer, Rome at
a gallop: Livy, on not gazing, jumping, or toppling into the void, in Larmour and Spencer (n. 4),
61101; see also Edwards (n. 4), 43.
11 Tac. Hist. 1.41.2, with the discussion in C. Damon, Tacitus. Histories I (Cambridge, 2003),
1834; she points out that Tacitus here exploits (as he does elsewhere) the symbolic power of the
event and location; see also Edwards (n. 4), 77.
12 On this temple in general, see I. Nielsen and B. Poulsen (edd.), The Temple of Castor and
Pollux: The Pre-Augustan Temple Phases with Related Decorative Elements (Rome, 1992) and I.
Nielsen, s.v. Castor, aedes, templum, LTUR 1.2425. N.W. DeWitt calls the area of the Forum
containing the Temple of Castor the plebeian end (Litigation in the Forum in Ciceros time,
CPh 21 [1926], 21824, at 221).
13 N. Purcell, s.v. Forum Romanum, LTUR 3.32536; M. Bonnefond-Coudry, Le Snat de la
Rpublique Romaine: De La Guerre dHannibal Auguste: Pratiques Dlibratives et Prise de
Dcision (Rome, 1989), 8090.
10

170

GEOFFREY S. SUMI

point of public meetings in the early period had always been the Curia and the
Comitium. In fact, the original Rostra was probably built on to the Curia and faced
the Comitium,14 thus serving as a platform from which one could address both people
and Senate; also, it could have been used by senators to inform the people of what
they had debated and decided in their meetings. After 338 B.C. a new Rostra was built,
still in the Curia/Comitium corner of the Forum but now between the Comitium and
the Forum. At some point in the next century orators began to turn away from the
Curia/Comitium complex to address the people who gathered in the Forum for
contiones.15 Elections continued to be held in the Comitium.
In the second century, the location of comitia changed as well: C. Licinius Crassus
(tr. pl. 145 B.C.), during one assembly of the people for the purpose of passing legislation, led voters into the Forum proper and perhaps to the Temple of Castor.16
Recent excavations have unearthed evidence to indicate that modifications were made
to the temple some time after 200 B.C. It is possible that this rebuilding accommodated
the new function of the temple.17 This combination of factors helped change the
orientation of the Forum, away from the Curia/Comitium and toward the Temple of
Castor.18
When L. Caecilius Metellus Delmaticus renovated the temple in 117 B.C., thus
increasing the size of the tribunal,19 which was now called a second rostra, he took
advantage of or at any rate acknowledged the place of the temple in the new orientation of the Forum. Raised platforms (pontes) provided access to and egress from the
tribunal, allowing citizens to cast their votes. The consequence of these modifications
to the temple was that it became a more important site for contiones and comitia.20
The final piece of the puzzle may have been the construction of the Gradus Aurelii
or Tribunal Aurelium (c. 74 B.C.). The precise location of this structure cannot be
determined with certainty.21 If it was near the Temple of Castor, which is a possible
inference from some of Ciceros references to it,22 it might have provided steps for the
people to stand on in front of the temple, while they attended a contio or awaited their
turn to vote in an assembly. This combination of the temple, its tribunal (a second
14 Perhaps on the steps leading from the Curia down into the Comitium, as L.R. Taylor
suggests (Roman Voting Assemblies [Ann Arbor, 1966], 223). Cf. also L. Richardson, Jr., A New
Topographical Dictionary of Ancient Rome (Baltimore, 1992), 3345; F. Coarelli, s.v. Rostra (Et
Repubblicana), LTUR 4.21213. On the development of the Curia/Comitium complex during
the Republic, see J.C. Anderson, The Historical Topography of the Imperial Fora (Brussels, 1984),
1220.
15 Plutarch claims that C. Gracchus was the first to do this (C. Gracch. 26.4).
16 Varro, Rust. 1.2.9; Cic. Lael. 96; Taylor (n. 14), 215.
17 Nielsen (n. 12), 1.243.
18 We should also note that already around 142 B.C. P. Scipio Aemilianus delivered a speech in
front of the temple, but the subject and occasion of this speech remain unknown (Festus 362L).
19 The tribunal of Temple 1A was c. 11 m wide and 6.50 m in depth (I. Nielsen and B. Poulsen,
The rebuilding of the first temple (temple IA), in Nielsen and Poulsen [n. 12], 806, at 84); the
tribunal of the Metellan temple was c. 21 m wide and 7 m deep (Nielsen and Poulsen, The Metellan temple, in Nielsen and Poulsen [n. 12], 87117, at 113). On the Metellan rebuilding, see also
Nielsen (n. 12), 1.244.
20 Taylor (n. 14), 219.
21 K. Korhonen, s.v. Tribunal Aurelium, Gradus Aurelii, LTUR 5.867, succinctly explains
the critical issues surrounding this structure. See also F. Coarelli, Il Foro Romano, 2 vols (Rome,
1992), 2.1909; Richardson (n. 14), 1812, 4001, believes the Tribunal Aurelium and Gradus
Aurelii were separate structures.
22 Cic. Dom. 54, Red. pop. 13, Sest. 34, Pis. 11. In these passages he refers to the structure as
Tribunal Aurelium. Earlier references (Clu. 93 and Flac. 66) call it the Gradus Aurelii.

MONUMENTS AND MEMORY

171

rostra), and the Gradus Aurelii would then have corresponded topographically to the
Curia/Comitium/Rostra complex diagonally opposite. The precise location of the
Gradus Aurelii notwithstanding, the rostra of the Temple of Castor provided an
effective vantage point from which a popular politician could gesture angrily at the
Curia across the Forum the symbol of a remote and insensitive ruling elite; and the
crowd would stand with its back to the Curia, instead of facing it, as it would if it
were listening to a speaker standing on the Rostra.23 The temple became the ideal
location for a champion of the people (such as P. Clodius Pulcher, tr. pl. 58 B.C. [see
further below]) to emphasize the gulf that existed between the senatorial aristocracy
and the urban plebs.
There can be no doubt that the changing topography of the Forum contributed to
the importance of the Temple of Castor in the political life of the city in the late
Republic. In the last generations of the Republic in particular, as one location for
contiones and comitia, the temple became a locus for popular politics and therefore
resonant of popular sovereignty. Controversial legislation provoked the fiercest
disputes in these often contentious assemblies.24 Two examples can illustrate what I
mean: Q. Caecilius Metellus Nepos (tr. pl. 62 B.C.) proposed a law that would have
conferred a command on Pompey to protect the city following the execution of those
involved in the Catilinarian conspiracy. M. Porcius Cato, who was also tribune in this
year, opposed it. When the people were gathering to vote on this law, Metellus and
Caesar surrounded the Temple of Castor with armed men and gladiators and perched
themselves atop the rostra of the temple. Cato, undaunted, ascended the tribunal and,
much to everyones astonishment, sat down between Metellus and Caesar. Metellus
began to read the text of the law that was being proposed, but Cato prevented him
from doing so by snatching the text from his hand. When Metellus began reciting the
law by heart, Q. Minucius Thermus, Catos friend, clapped his hand over Metellus
mouth. The crowd erupted; Metellus dispersed it and Cato was led into the temple.
The law was never passed (Plut. Cat. Min. 279). A few years later, during Caesars
consulship, Caesar was at the temple presiding over the assembly that would vote on
his agrarian law. His colleague in the consulship, M. Calpurnius Bibulus, entered the
Forum with his supporters to break up the assembly and prevent a vote. Some in the
crowd already assembled prevented Bibulus from ascending the rostra of the temple,
threw him down, snatched the fasces from his lictors and broke them, and (in one
account) dumped a basket of dung on Bibulus head. He retreated and the law
passed.25
23 Cicero describes the Curia as a sentinel that oversees the Rostra and stands ready to rein in
revolutionary oratory, thus implying that the Rostra was in the control of the senatorial aristocracy (Flac. 57): Hic, in hac grauissima et moderatissima ciuitate, cum est forum plenum iudiciorum,
plenum magistratuum, plenum optimorum uirorum et ciuium, cum speculatur atque obsidet rostra
uindex temeritatis et moderatrix offici curia, tamen quantos fluctus excitari contionum uidetis! Cf.
DeWitt (n. 12), 2201, who describes the area around the Rostra as having an aristocratic
character.
24 See e.g., Plut. Sull. 8.58, which relates the dispute over a law transferring the prestigious
command against Mithridates from Sulla (cos. 88 B.C.) to his rival Marius. To thwart this legislation, the consuls Sulla and Q. Pompeius Rufus, declared a suspension of public business. While
they were holding an assembly (probably a contio, Plutarchs language is unclear on this point) at
the Temple of Castor, Sulpicius and his supporters entered the Forum and routed them from the
temple, killing Pompeius son in the mle and forcing Sulla to flee. The next year another controversial law (distributing new citizens throughout all the tribes) provoked similar protests at the
temple (App. B Civ. 1.64.2902).
25 App. B Civ. 2.11.39; Cass. Dio 38.6.23; Plut. Cat. Min. 32.3.

172

GEOFFREY S. SUMI

Many politicians in the late Republic endeavoured to control the temple in their
hopes of swaying the assemblies that gathered there. It is not surprising, then, that the
politician who made the most concerted effort to be the peoples champion has been
accused of using the Aedes Castoris as his own armed camp. Cicero claimed that
P. Clodius Pulcher removed the steps of the temple to prevent access (for example,
Sest. 34), which can only mean that Clodius made it impossible for the temple to be
used as a site for comitia, perhaps by removing the pontes or otherwise blocking access
to them. Clodius had already conflated the usually separate functions of the comitia
and contiones. The traditional function of the contio was to present to the people the
merits of a particular bill (rogatio), but the people were not expected to respond. It
was not until the comitia that that they were formally asked (rogare) what they felt and
their vote was their response.26 Clodius, on the other hand, used contiones to put
questions to the people and allowed them to give a response. Cicero objected that
these responses came only from those who were members of Clodius gangs (operae)
and therefore did not constitute the will of the people. When Clodius prevented access
to the Temple of Castor, he (in Ciceros view) further impeded the traditional political
process in Rome. Clodius himself might have claimed that he was providing another
venue for the expression of popular sentiment.
In the political struggle following Caesars assassination the temple was the site of
contiones in which the major players in this struggle could address the urban populace
and Caesars veteran soldiers, whose allegiance was decisive to the eventual victor.
First, Cicero refers to a contio that M. Antonius held here either in June or November
of 44 B.C.27 He held another contio in October of 44 B.C. in which he accused Cicero of
being responsible for Caesars death (Cic. Fam. 12.3.2). This contio, coupled with a
new statue in Caesars honour, with the inscription To our father who well deserved
it, made clear the breach that divided Antonius from the conspirators and their
supporters. Next, Octavian returned to Rome in November of 44 B.C. at the head of
two legions of soldiers and in a contio swore an oath of his hopes to rise to his fathers
honours.28 In the case of this contio, Octavian might have been drawn to the Temple of
Castor by its proximity to the monument erected in Caesars honour, on which he
presumably swore his oath. It is worth bearing in mind that Octavians contio at the
temple on this occasion was a kind of ceremony of succession, inasmuch as he swore
to take his fathers place. This might have influenced Augustus incorporation of the
temple in his own ideology (see further below). Finally, Cicero mentions a statue of
L. Antonius (tr. pl. 44 B.C. and brother of the triumvir) that was erected near the
temple with a dedication describing him as patron of the voting tribes (quinque et
triginta tribus patrono, Cic. Phil. 6.12). Ciceros remarks about this statue are contained in a speech he delivered in 43 B.C., which shows that even in the triumviral
period the temple remained an important location for the expression of popular
sovereignty.
This brief survey of the changing topography of the Forum Romanum, which has
featured the political significance of the Temple of Castor, especially as a location for
popular assemblies in the late Republic, provides a basis for understanding why this
26

Taylor (n. 14), 28.

Cic. Phil. 3.27, 5.21; for an argument for the earlier date, see G.S. Sumi, Ceremony and
Power: Performing Politics in Rome Between Republic and Empire (Ann Arbor, 2005), 13841; for
the later date, see H. Frisch, Ciceros Fight for the Republic: The Historical Background of the
Philippics (Copenhagen, 1946), 151.
28 Cic. Att. 16.15.3; cf. App. B Civ. 3.41.169 and Cass. Dio 45.12.45.
27

MONUMENTS AND MEMORY

173

temple became symbolic of popular sovereignty in this period. It was a space over
which politicians seemed eager to exercise control. It is not surprising, then, that
Augustus, too, would want to demonstrate his control over this important monument
and its immediate vicinity. The advent of the Augustan Principate effected a change in
the political process. Popular assemblies, a hallmark of the Republican constitution,
were still held, but only under the princeps supervision. How decisive was Augustus
influence on the assemblies in his Principate is still unclear. Augustus actively
campaigned for his favoured candidates before they stood for election in the new
Saepta Julia in the Campus Martius (Suet. Aug. 56.1). Augustus might have wanted
elections to maintain an air of freedom,29 but it is likely that his own candidates
enjoyed a decided advantage. That elections were held in a structure named for the
Julian gens advertised who was the ultimate authority in the political process.
In a similar fashion, legislative assemblies came under the de facto control of the
princeps after 23 B.C. by virtue of his tribunician power, which enabled him to bring
legislation before the people, and his influence over the elections of consuls who
tended to be the other sponsors of legislation.30 These assemblies still met in the
Forum, but we cannot be sure of the precise location throughout the Augustan
Principate. We have evidence of a law passed in 9 B.C. at the new Aedes Divi Julii,
apparently in place of the Temple of Castor.31 Was this change of venue significant?
One explanation is that the Temple of Castor might have been unavailable due to
damage done by a fire a few years previously (14 B.C.). It was this damage that
prompted the Tiberian renovations that resulted in the rededication of the temple in
A.D. 6. We do not know whether all new legislation throughout the Principate was
voted on at this new location. If Augustus did decide to move legislative assemblies to
the Temple of Deified Julius, it is fitting that he chose a monument located in that part
of the Forum of symbolic significance to the Roman plebs. What is more, he chose a
Julian monument to demonstrate, in much the same way as he could for electoral
assemblies in the Saepta Julia, that in the Principate the ultimate authority in the
political process was the princeps and his family. The change from Aedes Castoris to
Aedes Divi Julii was, topographically, a small one; a slight turn to the left for the
citizen body which had been convened to vote on legislation. But symbolically it was
much more significant as legislative assemblies were further brought under the aegis
of the princeps and his family. The change of venue might show that Augustus wanted
to shift his focus from the more recent memories of the temples role in popular
politics in the late Republic, memories that recalled civil conflict, and embrace instead
more remote memories that could be exploited more effectively for his new ideology.32
In other words, the larger historical and mythological background for the Aedes
Castoris must be brought to account in considering the temples role in the formation
of Augustan ideology.
29 On elections in the Principate, see A.H.M. Jones, The elections under Augustus, JRS 45
(1955), 921 = id., Studies in Roman Government and Law (Oxford, 1960), 2750 and R. FreiStolba, Untersuchungen zu den Wahlen in der rmischen Kaiserzeit (Zrich, 1967), 87129.
30 G. Rotondi, Leges Publicae Populi Romani (Milan, 1912), 44262; cf. Sumi (n. 27), 2347.
31 This was a law enacted by T. Quinctius Crispinus (cos. 9 B.C.); Frontin. Aq. 129; cf.
M.H. Crawford, Roman Statutes, 2 vols (London, 1996), no. 63, 2.795, 797; Sumi (n. 27), 2356.
32 The bloodshed, civil strife, and memories of battles fought, violent and brutal conflicts in
which Romans clashed against their fellow Romans, could never be forgotten, but the memories
of recent war could be placed into context by using events from Romes more distant past to
demonstrate the citys resilience (Rea [n. 3], 4).

174

GEOFFREY S. SUMI

IV. THE FOUNDATION MYTH OF THE AEDES CASTORIS

The most frequently attested memory associated with the Aedes Castoris was its
foundation myth, which not only explained the circumstances of the temples
founding but also was adapted and retold in other contexts throughout Roman
history. Most importantly for our purposes, this myth remained alive under the
Principate and was retold in particular in the tradition surrounding the death of Nero
Drusus, Augustus stepson and Tiberius younger brother. Drusus was once marked
out as one of Augustus possible successors before his sudden and premature death (in
9 B.C.) while on campaign in Germany ended these hopes. Dio informs us that while
on this campaign Drusus was met by a woman of unnatural size who foretold his
imminent death. Afterwards, two young men were seen riding through his camp,
whose appearance was one of several omens that confirmed the initial prophecy.33
The story of the two riders appearing in Drusus camp after his death seems to have
been adapted from the foundation myth of the Temple of Castor in Rome and is an
appropriate starting point for an understanding of how this Republican myth could
be made to serve the ideology surrounding the succession.
The traditional story of the Dioscuri in Rome is their appearance at the battle of
Lake Regillus, when the Roman commander and dictator A. Postumius Albus
challenged the forces of Tusculum under the leadership of Octavius Mamilius. The
tyrant Tarquinius Superbus, recently driven from Rome, joined Mamilius army.
Postumius vowed to construct a temple to Castor in Rome if he was victorious in this
battle (Livy 2.20.12). In the heat of the contest, two young men, handsome and
unusually tall, appeared on horseback and led the Roman army to victory. Later,
these same two figures were seen in Rome watering their sweaty horses at the Lacus
Juturnae, just to the east of where the temple now stands, and brought news of the
successful battle.34 Just as important as the epiphany of the gods in battle was their
later presence in Rome where they announced the victory to the awaiting citizenry, in
this instance, a victory of freedom over tyranny. Already in Greek thought the
Dioscuri were often seen as saviour gods who provided timely assistance in battle,
leading armies from the brink of defeat to the glory of victory.35 The Dioscuri,
therefore, came to be associated with victories in battle and the swift announcement
33 The story of the two riders in Drusus camp appears only in Cass. Dio 55.1.35; cf. Suet.
Claud. 1.2 for the story about Drusus confrontation with the woman of unusual size.
34 Cic. Nat. D. 2.6; cf. 3.1113; Dion. Hal. 6.13.13; Val. Max. 1.8.1; Plut. Cor. 3.4; on the
announcement of the victory at Rome, cf. Plut. Aem. 25.2. See also Walter (n. 4), 1468. Livys
account of the battle (2.1920) does not mention the epiphany of the deities. However, he does
mention that the dictator Postumius established a reward to be given to the first two men to
penetrate the enemy camp (2.20.12). This detail might be Livys way of implicitly explaining away
the alternate tradition, of which he must have been aware, that related the epiphany of Castor
and Pollux: the two young men who appeared at the head of the Roman army were in fact these
soldiers who first penetrated the enemy camp. I know of no convincing explanation for the
absence of the Dioscuri in Livys account. Perhaps he was reluctant to attribute to divine intervention such an important victory in Roman history; D.S. Levene, Religion in Livy (Leiden,
1993), 153.
35 The Dioscuri were known as sotres from at least the fifth century B.C. They were credited
with aiding Lysander at the battle of Aegospotami (405 B.C.) (Plut. Lys. 12.1). For other
examples, see B. Poulsen, Cult, myth and politics, in Nielsen and Poulsen (n. 12), 4653, at 48,
and W.K. Pritchett, The Greek State at War. Part III: Religion (Berkeley, 1979), 1317, 212, 26,
27, 28. B. Poulsen argues that Augustus might have been inspired to adopt the Dioscuri as part of
his ideology by Hellenistic rulers (The Dioscuri and ruler ideology, SO 66 [1991], 11946, at
13742).

MONUMENTS AND MEMORY

175

of these victories often far from where the battles took place. They were at once
saviour gods and heralds of victory.36
This foundation myth, a favourite in the Roman historiographical tradition, was
adapted to suit the military campaigns of Romes imperial expansion. T. Quinctius
Flamininus, for instance, following his victory over Philip V at Cynoscephalae and his
later declaration of the freedom of Greeks (196 B.C.), dedicated silver shields to the
Dioscuri in the Temple of Apollo at Delphi,37 after which he was given the honorific
title soter,38 as the saviour and liberator of Greeks from the expansionary aims of
Philip V. Moreover, a coin from 126 B.C., minted by one of Flamininus descendants,
shows the Dioscuri standing on either side of a Macedonian shield a likely allusion
to Flamininus dedication.39 The Dioscuri had already been connected in Roman
history with a victory of freedom over tyranny at the battle of Lake Regillus. Flamininus adapted and broadened the association of these deities with victory and freedom
to include wars outside Italy, which influenced subsequent versions of the temples
foundation myth. An epiphany of the deities was said to have occurred at the battle of
Pydna when L. Aemilius Paullus defeated Perseus; again two young men appeared on
horseback to announce the victory and then later they were seen watering their horses
at the Lacus Juturnae.40 In 101 B.C. two young men on white horses made an appearance at the battle of Vercellae and announced in Rome the victory of C. Marius over
the Cimbri.41 In Florus version of this retelling of the legend, the two young men
delivered the letter announcing the victory (laureatae litterae) to the praetor in
front of the Temple of Castor.42 Thus, the Dioscuri were shown to approve of Romes
imperial ambitions.
Just as this legend came to be connected with victories on foreign soil, as the
Roman Empire expanded, so too was it adapted to the civil wars of the late Republic.
As one of several prodigies heralding Caesars victory over Pompey at Pharsalus (in
48 B.C.), two young men appeared in Syria, to announce this victory on the very day
that the battle took place.43 In this particular retelling of the myth, the Dioscuri were
merely heralds of victory.
36 For the Dioscuris connection with victory, see the coin of c. 211208 B.C. (M.H. Crawford,
Roman Republican Coinage, 2 vols [London, 1974], no. 61/1).
37 Plut. Flam. 12.11.
38 Plut. Flam. 16.7.
39 Crawford (n. 36), no. 267/1.
40 Cic. Nat. D. 2.6; on this passage, see the discussion in A.S. Pease (ed.), De Natura Deorum, 2
vols (Cambridge, MA, 1955), 2.55260; cf. 3.1113; Val. Max. 1.8.1; Plin. HN 7.86; Plut. Aem.
24.23, 25.12; Flor. 1.28.1415.
41 Plin. HN 7.86 (on this passage, see the discussion in M. Beagon, The Elder Pliny on the
Human Animal: Natural History, Book 7 [Oxford, 2005], 26971); Flor. 1.38. The association of
Castor and Pollux with victory continued into the Principate. A painting of Castor and Pollux
was set up in Augustus Forum alongside one of Victoria and Alexander (Plin. HN 35.27).
42 There is some evidence to indicate that the praetor urbanus was especially important as an
officiant in cult activity. Since the legend of Castor and Pollux usually included the
announcement of victory in Rome, whoever was left in charge of the city would have received the
news. In Florus account cited above, the praetor received the letter of victory in front of the
temple. In Dionysius account of the battle of Lake Regillus, the one who was left in charge of the
city (
) the praefectus urbi or the praetor urbanus?
conducted a futile search for the two young men who had announced the victory in Rome
(6.13.2). A certain Asellio (pr. urb. 89 B.C.) was murdered while making an offering to Castor and
Pollux in front of their temple (App. B Civ. 1.54; cf. Livy Per. 74; Val. Max. 9.7.4). In a much later
period (A.D. 216), the cult of Castor and Pollux in Ostia was administered by the praetor urbanus
(L.R. Taylor, The Cults of Ostia [Bryn Mawr, 1913], 223).
43 Cass. Dio 41.61.4:

176

GEOFFREY S. SUMI

Thus far, we have focussed on historiography, perhaps based on an oral tradition,

as the medium through which the memory of the foundation myth of the Aedes
Castoris was recalled. Other media were possible, too; particularly noteworthy is the
coin minted around 96 B.C. by A. Postumius Albinus, which depicts the Dioscuri
watering their horses at the Lacus Juturnae, a clear allusion to the temples foundation
myth.44 The minting of this coin is a reminder that particular gentes were often
responsible for preserving the memories of importance to their members. In this case,
the legend of the Dioscuri recalled the vowing of the temple by the minters distant
antecedent, A. Postumius Albus Regillensis, and its building later by his son, perhaps
in an effort to reclaim for the Postumian gens a myth that had been frequently retold
in the context of other Roman victories. The frequency with which this foundation
myth was retold and the number of Roman commanders who adapted it suggests that
it was no longer the exclusive property of the Postumian gens and had become a myth
for all Romans a national myth.
We now can begin to see the process whereby the foundation myth of the Aedes
Castoris was preserved, adapted and retold, first perhaps by members of the Postumian
gens (although evidence for this is lacking until the coin c. 96 B.C.) and later by other
Romans who desired a connection with the Dioscuri. The story of the Dioscuris
appearance following the death of Drusus is one piece of evidence for the further
adaptation of the story to suit Augustan ideology. A second piece is the story of
another mysterious and sudden appearance of two young men.
In Suetonius account of Caesars funeral, a divided crowd of mourners quarrelled
over the appropriate place for the cremation of Caesars body, with one group favouring the cella of the Temple of Capitoline Jupiter and another the curia of Pompeys
theatre, where Caesar had been assassinated. Suddenly two figures appeared, each one
girded with a sword and carrying two spears, and set fire to the bier with blazing
torches. Immediately a crowd of bystanders piled up branches, benches, seats and
other material.45 The dispute over the location of Caesars burial occurs in other
sources for his funeral,46 but Suetonius is our only authority for the dispute being
settled by the sudden appearance of two mysterious figures. The tradition of the
Dioscuris appearances in the Roman Forum, usually to announce a military victory,
and the description of these figures as armed warriors, are persuasive evidence for
understanding them to be the Dioscuri.47 Two questions in particular arise in connection with this story. First, can we determine its origin? And second, how does it
promote Augustan ideology?
The description of the two figures directing the crowd in the Forum Romanum, as
we have already noted, is somewhat reminiscent of Castor and Pollux directing the
battle line at Lake Regillus, yet at Caesars funeral they did so in a part of the Forum
that was associated with popular politics. One can understand, then, the Dioscuri in
the role of mob leaders in the minds of Romans who had become accustomed both to
. Dio recounts this epiphany in a long list of prodigies telling
of the outcome of the battle.
44
Crawford (n. 36), no. 335/10a and b.
45
Lectum pro rostris in forum magistratus et honoribus functi detulerunt. Quem cum pars in
Capitolini Iovis cella cremare pars in curia Pompei destinaret, repente duo quidam gladiis succincti
ac bina iacula gestantes ardentibus cereis succenderunt confestimque circumstantium turba uirgulta
arida et cum subselliis tribunalia, quicquid praeterea ad donum aderat, congessit (Suet. Iul. 84.3).
46 App. B Civ. 2.148; Cass. Dio 44.50.23; cf. Plut. Caes. 68.1, who describes the impromptu
cremation in the Forum but says nothing about a dispute.
47 The Dioscuri are consistently depicted on coinage carrying two spears each; see e.g.,
Crawford (n. 36), no. 44/5 (c. 211 B.C.).

MONUMENTS AND MEMORY

177

their sudden appearances in times of crisis and to their temple, which was at the centre
of the political turmoil of the late Republic. Suetonius account of the Dioscuris
appearance at Caesars funeral seems to be a conflation of memories, recalling both
the temples foundation myth as well as its role in late Republican popular politics.
If these figures were to be understood as the Dioscuri, however, we should acknowledge a change in their iconography from warriors on horseback or leading their
horses to ones armed with sword and spears.48 A possible explanation for this change
in iconography is the Dioscuris frequent identification with the Di Penates or Penates
publici. Much about this identification remains uncertain, but it is worth exploring
briefly what we do know. The Penates (sometimes Penates priuati) were deities who
protected and ensured the abundance of the household cupboard (penus) and were
worshipped in close connection with Vesta, the goddess of the hearth, and the Lar
Familiaris.49 The Penates publici had a temple on the Velia, which was purported to
have stood on the site of king Tullus Hostilius home.50 The royal home, then, became
the site of the public cult, an indication that the Penates publici protected the wellbeing of the community, along with Vesta, the goddess of the civic hearth.51 The
proximity of the Regia and the Temple of Vesta in the southeastern section of the
Forum is another manifestation of the connection of the royal household and civic
hearth.52 These structures may have been part of a larger complex of buildings
devoted to the Roman kingship.53 The temple on the Velia is first mentioned by Livy
(45.16.5) in the context of omens that appeared in the year 167 B.C., although it is
probably much older.54 Dubourdieu has argued that the identification of the Dioscuri
with the Penates publici was probably late, as the coins of M. Fonteius demonstrate
(minted in 108 or 107 B.C.).55 More to the point, for our purposes, is the relief on the
Ara Pacis, showing Aeneas sacrificing at Lavinium, where one can see the shrine of
the Penates in the background,56 in which they are depicted as twin gods, seated, and
holding spears,57 armed in a fashion similar to the Dioscuri but not as mounted
warriors. This relief demonstrates that the Penates became part of the Aeneas legend
certainly in the Augustan period but some connection with Troy is in evidence before
then.58 What we have in Suetonius account of Caesars funeral, then, is the epiphany
48 For the depiction of the Dioscuri as mounted warriors, see e.g., Crawford (n. 36), no. 44/5
(there are numerous other examples); for a depiction of them leading their horses, see e.g.,
Crawford (n. 36), no. 304/1.
49 J. Linderski, s.v. Penates, Brills New Pauly (LeidenBoston, 2007), 10.71820.
50 Varro, ap. Non. 852L; Solin. 1.22; Richardson (n. 14), 289. Augustus includes this temple in
a list of those that he built (feci, Mon. Anc. 19).
51 Livy frequently uses the formula penates publici priuatique in a manner that makes them
synonymous with ones homeland (patria) (3.17.11, 22.1.6, 25.18.10, 45.24.12).
52 A. Alfldi, Early Rome and the Latins (Ann Arbor, 1971), 258.
53 This is the theory of Coarelli (n. 21), 1.5678, which still has many detractors. For a critique,
see F.E. Brown, Gnomon 56 (1984), 3813. A summary of modern theories about the Regia can
be found in T.J. Cornell, The Beginnings of Rome: Italy and Rome from the Bronze Age to the
Punic Wars (c. 1000264 BC) (London, 1995), 23941. The Regia was rebuilt by Cn. Domitius
Calvinus in 36 B.C. (Cass. Dio 48.42.46; Plin. HN 34.48; CIL 6.1301=ILS 42).
54 A. Dubourdieu, Les Origines et le Dveloppement du Culte des Pnates Rome (Rome,
1989), 4401.
55 Dubourdieu (n. 54), 4309; Crawford (n. 36), no. 307/1a.
56 Serv. ad Aen. 3.12 (quoting Varro); S. Weinstock, Two archaic inscriptions from Latium,
JRS 50 (1960), 11218, at 113.
57 As Dionysius describes them on the temple on the Velia (1.68.1).
58 When the idea of Trojan ancestry became established in Rome, probably around the middle
of the 4th cent. BC, the sacred objects associated with Troy, such as the Palladium, were counted
among the Penates (Linderski [n. 49], 719).

178

GEOFFREY S. SUMI

of twin gods in the Forum directing a crowd of Romans, in the manner of the Dioscuri, but armed like the Penates publici, to cremate Caesars body near monuments
connected with kingship and the public hearth. To identify these figures as either the
Dioscuri or the Penates publici is futile. They had characteristics of both at once.
The question remains as to the source of this story peculiar to Suetonius. Certainty
is impossible, but one hypothesis is worth exploring. Of crucial importance is the role
first of the crowd at Caesars funeral in determining (with the help of these two
figures) a location for the cremation of the body, and later marking the location with a
monument in Caesars memory. The crowd chose a location directly in front of the
Regia,59 which served as the official headquarters of the Pontifex Maximus, Caesar
himself in 44 B.C., and behind which was the home of the Pontifex Maximus, the
Domus Publica. Moreover, as Pontifex Maximus, Caesar was chief priest of the public
cult, which found its focus in the public hearth represented by the Temple of Vesta
adjacent. The association of the Penates publici with Vesta, and Vesta with the Regia,
might have encouraged the conflation of the Dioscuri and Penates publici at the basis
of Suetonius account.
The first phase in commemorating the location of Caesars cremation was the
erection of a monument (variously identified as an altar or column),60 which came to
serve as a rallying point for Caesars supporters and acquired cult status. If anyone
ventured to wonder why Caesars body was cremated where it had been, and therefore
why the monument stood where it did, this story provided divine sanction for the
crowds choice of the location in the Forum. Moreover, this story linked Caesars
monument closely with Roman kingship. We should bear in mind that it was the
urban crowd that hailed Caesar as king when he returned to Rome in January of 44
B.C., after celebrating the Feriae Latinae on the Alban Mount, and was seen wearing
the high red boots of the Alban kings. Caesar wittily rejected the crowds
pronouncement by reminding them, as if they had mistaken his cognomen, that he was
not Rex but Caesar. In the month following at the Lupercalia, M. Antonius offered
Caesar a diadem by order of the people.61 The dictator, seated on the Rostra
resplendent in purple, refused it to the delight of the assembled crowd, and then
pointedly sent the diadem to the Temple of Capitoline Jupiter, recording the act in the
Fasti and declaring that only Jupiter was king of the Romans (Cass. Dio 44.11.3).
The second phase of commemoration of the location of Caesars cremation was
59 Suetonius account does not include this detail, but we have the testimony of Appian, who
states that the crowd cremated Caesars body near the palace (basileion), presumably a reference
to the Regia (B Civ. 2.148.616). We can also infer it from topographical evidence (see further next
note).
60
Since most of our ancient sources for Caesars funeral do not describe specifically where
Caesars body was cremated, we must infer the location from the future site of the monument that
the plebs erected in his honour (Cass. Dio 47.18.4). The rostra of the Temple of Divus Julius
contained a niche to accommodate this monument (R. Ulrich, The Roman Orator and the Sacred
Stage: The Roman Templum Rostratum [Brussels, 1994], 1812). Livy states that Caesars body
was cremated ante rostra, which, if he means the Rostra at the west end of the Forum, is an error;
but he could mean the rostra on the Temple of Castor, in which case he also connected the
cremation closely with the temple (Per. 116). Appian (as noted above, n. 59) claims that Caesars
cremation took place near the ancient palace of the kings of Rome (a probable reference to the
Regia). This monument, of course, was in the same part of the Forum, not far from the Temple
of Castor, but with his allusion to the Regia Appian seems to want to connect Caesars cremation
with ancient Roman kingship.
61 populi iussu: Cic. Phil. 2.87; cf. Cass. Dio 44.11.2. Other sources are Suet. Iul. 79.2; Plut.
Caes. 61.

MONUMENTS AND MEMORY

179

the erection of the Aedes Divi Julii by Octavian in 29 B.C. on the spot of Caesars
cremation and incorporating the memorial erected in Caesars honour. Of all the
rebuilding of Roman monuments that Augustus undertook during his reign, he only
constructed four new temples: those of Apollo Palatinus, Divus Julius, Jupiter Tonans
and Mars Ultor. A recent study has pointed out that of the four only the latter was
a manubial building, constructed after a vow taken by Octavian at the battle of
Philippi. Augustus erected the Temples of Jupiter Tonans and Apollo Palatinus at the
site of lightning strikes, clear signs from the gods as to the appropriate location for his
new temples.62 We can reasonably believe that Augustus desired a similar sign from
the gods for the location of his temple to the Deified Julius, which he found in this
story that Suetonius recounts. That the story also gives the Roman people a role to
play in the founding of his temple was an additional attraction.
There are other reasons to believe that Augustus would have been drawn to such a
story. First, the Penates publici recalled the Aeneas legend, of seminal importance to
the Julian gens and one that Augustus elevated to a national myth.63 Second, the
deities who later became identified with Augustus successors were shown here at the
beginning of the dynasty orchestrating the cremation of Caesars body, the first step
in his apotheosis, which cleared the way for Augustus to succeed (initially as Diui
filius). Finally, the Penates publici recalled the public cult overseen by the Pontifex
Maximus, Caesar himself at his death and Augustus beginning in 12 B.C. The story
thus drew a direct link between the son succeeding his father as Chief Priest, skipping
over the tenure of the inconveniently long-lived M. Aemilius Lepidus.
One can see how the story of the two figures at Caesars funeral could develop from
the Dioscuris traditional mythology which recalled their sudden and timely
appearances at moments of crisis. The aspect of Suetonius account that is not in
accord with this same traditional mythology and their identification with the Penates
publici, namely the Dioscuris depiction as leaders of a crowd in the Forum instead of
an army on the battlefield, can be explained as a product of the political function of
their temple. This story, a modification of the Dioscuris traditional mythology, gave
these gods a role to play in the transformation of Caesar from mortal to divine. Their
appearance at Drusus death might have had the same significance.
V. THE TRANSVECTIO EQUITUM AND
THE SUCCESSION
The appearance of the Dioscuri at Drusus camp suggests a connection between the
Dioscuri and Augustus successors.64 More direct evidence for such a connection is
the transuectio equitum, an annual parade involving members of the equestrian order
who had the right of the public horse (equites equo publico). The parade began at the
Temple of Mars on the Via Appia and wound its way to the Forum, where it ended in
front of the Temple of Castor. This parade took place on 15 July each year and,
according to one tradition, commemorated the role of the Dioscuri in the victory at
62 O. Hekster and J. Rich, Octavian and the thunderbolt: the temple of Apollo Palatinus and
Roman traditions of temple building, CQ 56 (2006), 14968, at 153.
63 See A. Erskine, Troy between Greece and Rome: Local Tradition and Imperial Power (Oxford,
2001), 1543, for a discussion of this process and some of its implications.
64 Other scholars have noticed this connection; see in particular Poulsen (n. 35) and A.J.
Woodman, Mutiny and madness: Tacitus Annals 1.1649, Arethusa 39 (2006), 30329, esp.
30811.

180

GEOFFREY S. SUMI

Lake Regillus (in 499 or 496 B.C.).65 Thus, in the Augustan period, there was an annual
ceremony that cued Romans to recall the foundation myth of the Temple of Castor
in much the same way that Ciceros words at the Temple of Jupiter Stator evoked the
foundation myth of that temple, through which he was able to compare himself with
Romulus.
When Augustus revived this ceremony (after a long hiatus, according to Suetonius
[Aug. 38.3]), he apparently altered its emphasis and brought it securely under the aegis
of the imperial family by placing his heirs in prominent positions. In the Republic
equites equo publico paraded down to the Forum, each one leading his own horse,
where they presented themselves to the censors, seated at the Temple of Castor, who
discharged them from military service (Plut. Pomp. 22). Suetonius account of the
Augustan revival (Aug. 38.339) lays emphasis on this ceremony as a review (recognouit)
of knights who were required to give an account of their lives to the princeps assisted
by ten members of the Senate. The severest punishment seems to have been loss of
rank (ignominia notauit); others might be reprimanded for scandalous conduct.
Dionysius of Halicarnassus, writing under the Augustan Principate and presumably a witness to the Augustan revival, describes a ceremony of great pageantry in
which Roman knights paraded on horseback, arrayed in tribes and divisions, crowned
with olive branches and wearing purple robes adorned with emblems of their valour
a spectacle, in Dionysius estimation, worthy of Romes empire. Furthermore,
Augustus grandsons, Gaius and Lucius Caesar, led the parade under the title
principes iuuentutis,66 apparently fitted out in a manner reminiscent of the Dioscuri.
By giving his successors a leadership position in this ceremony he further advertised
the role that they would play in overseeing the acquisition and maintenance of empire.
A Republican ceremony with a primarily censorial function thus became a rite of
succession that adumbrated a glorious future for the Roman Empire under the aegis
of Augustus successors.
A second tradition holds that this parade was instituted in 304 B.C. by the censor Q.
Fabius Maximus Rullianus.67 It is more likely that this tradition preserves the memory
of a revival of the ceremony than its institution, as well as its transformation from a
strictly religious ceremony to one with a censorial function. The increase of the
number of cavalry voting units (centuriae) from the original six to twelve might also
have occurred around this time.68 In the Augustan revival of this ceremony the partici65 Dion. Hal. 6.13.4; Plut. Cor. 3.5; cf. Suet. Aug. 38.3. On the parade in general, see S. Weinstock, Rmische Reiterparade, SMSR 12 (1937), 1024; on the significance of the parade in the
Augustan Principate, see Spencer (n. 10), 8997.
66 This title for C. and L. Caesar was widely advertised on inscriptions and coins; see e.g.,
V. Ehrenberg and A.H.M. Jones, Documents Illustrating the Reigns of Augustus and Tiberius
(Oxford, 1949), no. 65 (=AE 1899, 153) and RIC2 1, p. 55, no. 205. See also RIC2 1, p. 54, no. 198,
for a depiction of C. Caesar on a galloping horse armed with sword and shield. J.B. Ward-Perkins
points out that the equites continued to be associated with the emperors heirs, alluding in
particular to the decursio at imperial funerals, such as the one depicted on the base of the column
of Antoninus Pius (Columna Divi Antonini, in P. Ducrey (ed.), Mlanges dhistoire ancienne et
darchologie offerts Paul Collart [Lausanne, 1976] = H. Dodge and B. Ward-Perkins (edd.),
Marble in Antiquity: Collected Papers of J.B. Ward-Perkins, Archaeological Monographs of the
British School at Rome 6 [London, 1992], 10714, at 1078); cf. also L. Vogel, The Column of
Antoninus Pius (Cambridge, MA, 1973), 5668.
67 Livy 9.46.15; Val. Max. 2.2.9; Aur. Vict. Caes. 32. On the institution of the transuectio
equitum, see also S. Oakleys note on Livy 9.46.15 in A Commentary on Livy, Books VIX, 4 vols
(Oxford, 19972005), 3.6425.
68
A. Momigliano, Procum Patricium, JRS 56 (1966), 1624.

MONUMENTS AND MEMORY

181

pating equites were divided into six turmae, which were probably an allusion to the
original six centuries.69 This ceremony, and thus the temple, had close associations
with the equestrian order whose patron deities were Castor and Pollux.70 In the late
Republic, the equestrian order became increasingly politicized, first by C. Gracchus
who in his tribunate succeeded in identifying issues of particular importance to the
equestrian order and thereby separating its interests from those of the Senate. Central
to Ciceros rhetoric two generations later was concordia ordinum, his succinct
expression of hope for rapprochement between the orders. When Augustus came to
power, he began the process of separating career paths for members of the senatorial
aristocracy and equestrian order,71 thereby succeeding where Cicero had failed in
establishing a concordia ordinum. The revival of the transuectio equitum enabled
Augustus to advertise himself as patron of the equestrian order,72 just as he was
princeps senatus, and protector of the urban plebs (by virtue of his tribunician power).
Augustus successors, as principes iuuentutis, were shown to be leaders of the next
generation of equites.
The performance of the transuectio equitum in the Augustan period, then, was
more a glorification of the imperial family than a celebration of the victory at Lake
Regillus.73 We should not, however, dismiss the fact that Augustus chose to glorify the
imperial family through the identification of his successors with the Dioscuri in the
context of a ceremony that recalled a great victory from Romes remote past and was
of central importance to the equestrian order. The memories associated with the
Temple of Castor were the symbolic framework for Augustus successors and
ultimately shaped the way that Augustus communicated his ideology.

VI. DIOSCURI AS DEMIGODS: BETWEEN LIFE

AND DEATH
The Dioscuri were appropriate divine counterparts for Augustus successors also
because of their traditional mythology. Cicero listed Castor and Pollux, along with
Hercules, Liber, Aesculapius and Quirinus, as demigods heroes whose achievements
in their lifetimes were so grand that they were accorded divine status (Cic. Leg. 2.19).
Horace, writing a few decades later, imagines a scene with Augustus one day reclining with Pollux and Hercules, Dionysus and Quirinus.74 Further evidence for the
importance of the Dioscuri and other demigods in Augustan ideology was a heavily
69

Mommsen, Staatsrecht, 3.523.

M. Albert, Le Culte de Castor et Pollux en Italie (Paris, 1883), 819. Moreover, the Campani
equites received the right of Roman citizenship in 340 B.C., and a bronze inscription commemorating the event was set up at the Aedes Castoris (Livy 8.11.1516, with the discussion in Oakley
[n. 67], 2.51315). They may have also been given the equestrian census rank.
71 P.A. Brunt, Princeps and equites, JRS 73 (1983), 4275, at 434.
72 Augustus himself had been a member of an equestrian family before being adopted into the
patrician Julii (Suet. Aug. 2.1; cf. Iul. 41.1; Plut. Caes. 58.1; Cass. Dio 43.47.3). He evinced no
shame at his equestrian birth and even boasted of it in his own writings, remarking that his father
was the first member of his family to enter the Senate. In his rivalry with Antony, however,
Octavian had to endure many insults directed at his character and the relative obscurity of his
familys origin (Suet. Aug. 2.3).
73 As Weinstock concluded ([n. 65], 24).
74 Hor. Carm. 3.3.918; cf. Ep. 2.1.56, where he enumerates some of those gods who achieved
immortality through their great deeds (Romulus, Liber pater, Castor and Pollux). Later in the
same poem (l. 15) Horace compared these demigods, all of whom achieved their divine status
after their lifetimes, to Augustus who was praesens diuus (cf. Carm. 3.5.23).
70

182

GEOFFREY S. SUMI

frequented corner of the Forum Augustum dedicated to such figures, with a painting
of Castor and Pollux placed next to one of Alexander, another mortal who achieved
divine status, with Victory in between (Plin. HN 35.27). The traditional story of the
death of Castor as told by Pindar is pertinent here. After a quarrel with the sons of
Aphareus (over the division of cattle taken in a raid) Castor was killed, and Pollux, in
grief over his brothers death, prayed to Zeus that he might die with him. Zeus gave
him a choice of immortality for himself and death for Castor, or they both could live
on Olympus for one day and spend the next in Hades.75 (A variant version has one
living in heaven and the other in Hades on alternate days.)
The traditional mythology of the Dioscuri, which has them inhabiting both the
mortal and divine realms, is the foundation for modern scholars assertions that
associate the Dioscuri with death and the afterlife. For this very reason, E. Strong
regards them as emblems of future life.76 Their frequent depiction on sepulchral
reliefs was the result of their cult having developed out of that of the heroised dead
and they never lost their connection with the underworld.77 This belief might have
come to Rome from Hellenistic royal cult, since the Diegesis of Callimachus claims
that the poet described the sudden death of Arsino as her having been snatched away
by the Dioscuri.78 F. Cumont points out that the Dioscuris double existence was the
basis of the ancients view of them as representing the two hemispheres of the sky, the
regions of the sun and the moon, of light and darkness.79 Another aspect of the
Dioscuris traditional mythology has them protecting travellers, especially sailors, for
whom they ensure a happy voyage. Cumont posits a natural transformation of the
Dioscuri from protectors of travellers to protectors of the dead.80 Even if one does
not accept his cosmic interpretation of the Dioscuri as symbols of the two
hemispheres, the story of their alternate deaths and their achievement of immortality
through merit was memorable and compelling, sufficient to associate them with
funerals and death.81 This might help explain their appearance at Caesars funeral and
75 Pind. Nem. 10.4991. Horaces language in Ep. 2.1.56 (cum Castore Pollux recepti)
perhaps shows that he was thinking of the same story. Ovid retells it in the Fasti (5.693720).
76 E. Strong, Apotheosis and After Life (London, 1915), 201.
77 Strong (n. 76), n. 27 on p. 275. For their iconography in ancient art in general, see LIMC
3.1.567635; for their appearance on Roman sarcophagi, see G. Koch and H. Sichtermann,
Rmische Sarkophage (Munich, 1982), 144.
78 Dieg. 10.10 on Call. fr. 228 (Pfeiffer). On the Diegesis in general, see R. Pfeiffer (ed.),
Callimachus, 2 vols (Oxford, 1949/1953), 2.xxviii.
79
F. Cumont, Recherches sur le Symbolisme Funeraire des Romains (Paris, 1942), 64103,
citing in particular Sext. Emp. Math. 9.37 in n. 1 on p. 68. He is followed by J.M.C. Toynbee,
Death and Burial in the Roman World (Ithaca, NY, 1971), 38, who also regards the Dioscuri as
guardians of the dead (194); cf. M. Mackintosh, The Divine Rider in the Art of the Western
Roman Empire (Oxford, 1995), 38. Cumonts views on the symbolic significance of sepulchral art
remain at the heart of a vexed question among art historians, namely whether we should view
mythological scenes in sepulchral art as representative or symbolic of the ancients views of death
and the after life or as merely ornamental. See Nocks thorough review of Cumonts book
(Sarcophagi and symbolism [Review of Cumont, 1942], AJA 50 (1946), 14070). He points out
that the Dioscuri were often used as framing figures, placed on either side of another figure (e.g.
Helen or another goddess) (152, n. 48). More recently, R. Ling (A relief from Duke Street,
Aldgate, now in the Museum of London, Britannia 24 [1993], 712) reminds us that many
popular motifs of sepulchral art appear in other contexts as well. Clearly, some of Cumonts
statements go too far (e.g. the one that sees Pythagorean philosophical ideas in much tomb art
[p. 251]), but his interpretation in my view is generally sound.
80 Cumont (n. 79), 65.
81 As Nock puts it, any ordinary ancient, reading of them or seeing their familiar type,
would think rather of their human story, their alternate deaths, their attainment of heaven for
merit (n. 79), 151.

MONUMENTS AND MEMORY

183

also at Drusus camp at the time of his death. Moreover, the brotherly affection at the
heart of the traditional mythology of the Dioscuri was especially relevant for Drusus
and Tiberius, since in a magnificent display of fraternal pietas Tiberius famously
recovered Drusus body after his death and transported it back to Italy (discussed
further below).
Castor and Pollux once inhabited the mortal realm, and in death they spent time
both in the underworld of mortals and in the heaven of the gods. We should note, too,
their capacity to move comfortably between the mortal and divine realms, as well as
from the realm of death to that of life (hence their later function as funerary deities).
The mythology of Castor and Pollux underscored the notion of semi-divine further
by having them share their immortality, each one taking turns on Olympus. For this
reason, they came to be regarded as gods on earth. In light of this it is not surprising
that Augustus might want to use the Dioscuri as divine counterparts for members of
the imperial family who aspired to a similar status.
Furthermore, by identifying his successors with Castor and Pollux (the DIOSkouroi), Augustus was also identifying himself with Zeus/Jupiter, the king of the gods.
In general, Augustus seems to have avoided identifying himself directly with Jupiter,
choosing instead to present himself as Jupiters agent.82 In fact, scholars have
remarked on a topographical shift away from the Capitoline, the site of Jupiters
temple, and toward the Palatine and Augustus new temple of Apollo.83 Augustus
would still have been drawn to many of Jupiters attributes. For instance, Jupiter
secured and defended Rome by virtue of his temple on Romes citadel, now the capitol
of an expansive empire. Jupiter received vows as well as the spoils of war borne in
triumph. Moreover, he was connected with constitutional government, which
Augustus was at pains to restore following the civil wars.84
VII. THE DIOSCURI AND FRATERNAL HARMONY IN THE
IMPERIAL FAMILY
The identification of the Dioscuri with Augustus successors, already in evidence in
the transuectio equitum, occurred again in the tradition surrounding Tiberius and
Drusus as well as other members of the imperial family. We have already discussed the
story of the appearance of the Dioscuri in Drusus camp after his death, which is but
one connection between Augustus stepsons and the twin deities. Valerius Maximus
makes the connection more explicit in his account of Tiberius journey to recover
Drusus body after the latters death on military campaign. Tiberius exhibition of
fraternal devotion and pietas, Valerius tells us, was reminiscent of the Dioscuri.85 On
82 Suetonius relates a dream of Cicero who saw Augustus receiving a whip from Capitoline
Jupiter (Aug. 94.9; cf. Plut. Cic. 44; Cass. Dio 45.2); Rea (n. 3), 5661.
83 P. Zanker, The Power of Images in the Age of Augustus (Ann Arbor, 1988), 108; cf. Suet. Aug.
91.2; see also J.R. Fears, The cult of Jupiter and Roman imperial ideology, ANRW 2.17.1,
3141, at 601.
84 Mon. Anc. 34.12; Rea (n. 3), 57; Fears (n. 83), 65.
85 Val. Max. 5.5.3; the story of Tiberius recovery and transport of Drusus body was a famous
one; see also Livy Per. 142; Strabo 7.1.3; Plin. HN 7.84; Sen. Epist. 11.15.5; Suet. Tib. 7.3; Cass.
Dio 55.2.1. Cf. Ov. Fast. 1.7058 for an oblique reference connecting Tiberius and Drusus with
the Dioscuri. The story of Tiberius recovery of Drusus body was so important that Tiberius
may have commemorated it in the sculptural programme at his grotto in Sperlonga (R.G.M.
Nisbet, Notes on the text and interpretation of Juvenal, in N. Horsfall (ed.), Vir Bonus Discendi
Peritus. Studies in Celebration of Otto Skutschs Eightieth Birthday, BICS Supplement 51
(London, 1988), 105, n. 29. Tacitus uses the occasion of Germanicus funeral to set the honours

184

GEOFFREY S. SUMI

this occasion, Tiberius and Drusus could play the roles of Castor and Pollux, as the
one brother mourned the death of the other, in a manner somewhat reminiscent of the
Dioscuris traditional mythology (discussed above).
Valerius Maximus story also shows a further development of Augustan ideology,
by introducing the notion of fraternal harmony. This was especially important in the
first succession, since the early Principate featured brothers within the imperial family
who were marked out as possible successors. Castor and Pollux were regarded as a
paragon of fraternal harmony, as noted in the story that Valerius Maximus relates, in
contra-distinction to another pair of famous mythological brothers, Romulus and
Remus, whose relationship ended in fratricide, a symbolic bloodletting that infected
the Roman psyche and led to periods of civil war.86 The rededication of the Temple of
Castor in the name of Tiberius and his brother Drusus drew a close link between
Augustus stepsons and their divine counterparts. Even more telling is the renovation
of the Temple of Concordia in the Forum Romanum, which Tiberius undertook
following his triumph in 7 B.C. In a manner similar to the Aedes Castoris the Aedes
Concordiae was rededicated in Tiberius name and that of his brother Drusus on 16
January A.D. 10, the anniversary of Octavians adoption of the title, Augustus.87 The
coincidence of dates was probably intentional, so that Augustus could underscore the
association of this temple, rededicated close to the end of his own life and reign, with
transfer of power.
The Temple of Concordia, first founded according to tradition by Camillus in 367
B.C. amid the Struggle of the Orders and later refurbished by L. Opimius following the
death of C. Gracchus in 122 B.C., adumbrated the harmony within the city during the
most tumultuous times of the Roman Republic. This temple figured in much of the
strife of the late Republic, as the site of Ciceros Third Catilinarian and where M.
Antonius and Cicero traded barbs after Caesars assassination (Cic. Phil. 5.1820).
The new temple came to be known as the Temple of Concordia Augusta (according to
the Fasti Praenestini [Inscr. Ital. 13.2, p. 115]), which effectively redirected the
symbolic orientation of the temple from city to imperial family. Concordia, an
important symbol of the Republic, thus became subsumed under the ideology of the
imperial family. The message was clear: only concord within the imperial family could
ensure the peace and prosperity of the Roman world.88 Moreover, a coin minted by
Tiberius (c. A.D. 356) depicting the faade of this temple on its obverse shows a
pediment crowned by several figures (RIC2 1, p. 98, nos 55, 61 and 67). The central
three are apparently female figures, probably meant to represent Concordia, Pax and
Germanicus received against those of his father. In particular, he mentions the journey that
Germanicus brother (probably his adoptive brother, Drusus, Tiberius son) made to meet
Germanicus remains: Drusus deigned to go only as far as the city gates. Thus, Tacitus can
demonstrate how much has changed from Augustus regime to Tiberius, and in particular imply
the end of fraternal harmony (Ann. 3.5; cf. A.J. Woodman and R.H. Martin (edd.), The Annals of
Tacitus, Book 3 [Cambridge, 1996], 989).
86 These issues are discussed in C. Bannon, The Brothers of Romulus (Princeton, 1997),
17881.
87 The Fasti Praenestini record the temples natalis dies (16 January); CIL 1.1, p. 231; Inscr. Ital.
13.2, p. 115. See also, Ov. Fast. 1.63739. Cass. Dio 56.25.1 has A.D. 10 for the year of dedication.
On this temple in general, see A.M. Ferroni, s.v. Concordia, aedes, LTUR 1.31620; for its
decorative programme, see B.A. Kellum, The city adorned: programmatic display at the aedes
Concordiae Augustae, in K. Raaflaub and M. Toher (edd.), Between Republic and Empire: Interpretations of Augustus and His Principate (Berkeley, 1990), 276307.
88 This concept was underscored by the statues of Pax, Concordia and Salus that Romans
honoured.

MONUMENTS AND MEMORY

185

Salus. The flanking figures may represent Tiberius and Drusus, the dedicatees of this
temple, in the pose of warriors holding spears. It is possible that we are meant to think
of their identification with other spear-toting brothers, namely, the Dioscuri.89 I
would take this one step further and argue that the figures depicted are actually the
Dioscuri themselves. They are, after all, frequently depicted holding spears and used
as flanking figures in Roman art.90 If this interpretation is correct, then the Dioscuri
oversaw, along with the other deities depicted, the harmony of the imperial household
and ultimately had a hand in protecting the tranquil transition of power through
dynastic succession.
The two buildings under discussion both bore the names of Tiberius and Drusus.
Even if the flanking figures on the pediment were not the Dioscuri, other scholars
have noted a similarity in the architectural decoration of these two buildings, after the
Temple of Concordia was refurbished in A.D. 10.91 These two buildings, then, were
closely linked. The rededication of the Temple of Concordia in the names of Tiberius
and Drusus combined the notion of fraternal harmony within the imperial family
with the more traditional Republican concept of political or civic harmony.
The rededications of both temples, of course, took place after Drusus death,
when harmony between the two brothers was no longer necessary, or possible. Their
relationship, nonetheless, may have served as a model for other relationships within
the imperial household, most notably that of Germanicus and Tiberius son Drusus,
who were among the hierarchy of heirs before and after Augustus death. In A.D. 9,
following victories in Illyria, Tiberius, who was by this time heir designate, was
awarded a triumph, which he postponed because of Quinctilius Varus disastrous
defeat in Germany. Augustus, nonetheless, went out to meet Tiberius as he returned to
the city, accompanied him to the Saepta Julia and there, flanked by the two consuls on
a tribunal constructed for the purpose, princeps and heir greeted the people.
Germanicus made the announcement of victory and was awarded the lesser distinction of triumphal honours (ornamenta triumphalia). Tiberius son Drusus, although
he had played no part in the campaign, was given the privilege of attending meetings
of the Senate and of voting before the ex-praetors. On this occasion, Augustus conferred honours on members of his household in order of their proximity to the throne
first Tiberius, then Germanicus, and finally Drusus. Augustus used the public
conferral and announcement of this triumph as an opportunity to demonstrate the
stability of the Principate, recently imperilled by the disaster in the Teutoberg Forest,
by honouring three men in a grand ceremony of succession,92 which represents
Augustus best attempts to lay the groundwork for a harmonious and tranquil transfer
of power. The date of the rededication of the Temple of Concordia, as I already
noted, marked the anniversary of Augustus assumption of power; it also corresponded closely with this ceremony of succession and may have been another attempt
at reassuring the populace that harmony among the heirs in fact existed at this crucial
moment in the history of the Principate.
89 Ferroni (n. 87), 1.318. Mattingly (BMCRE 1.116, p. 137) and Sutherland (RIC2 1, p. 96)
identify the central figures as the Capitoline triad, Jupiter, Juno and Minerva. Mattingly
identifies the flanking figures as Ceres and Diana while Sutherland declines to attempt a specific
identification.
90 E.g. the famous painting of Apelles in the Forum Augustum depicting Alexander between
the Dioscuri (Plin. HN 35.93).
91 Kellum (n. 87), 276307, at 277; cf. Nielsen (n. 12), 1.245.
92 Suet. Tib. 17.2; Cass. Dio 56.17; cf. Vell. Pat. 2.121.23.

186

GEOFFREY S. SUMI

The mythical and divine models for these earthly relationships, as noted above,
were the Dioscuri, whose temple after A.D. 6 bore the names of Tiberius and Drusus.
These deities were bequeathed as divine models to two of the participants in
Augustus grand ceremony of succession, Germanicus and the younger Drusus.93
Tacitus also remarks on their harmonious relationship, in direct contrast to the
rivalry between those closest to them, including their wives.94 The Dioscuri continued
to be associated with the fraternal harmony of imperial heirs and hence with the
tranquil transition of power under dynastic succession.

VIII. CONCLUSION
A critical element in understanding the importance of the memories evoked by the
Aedes Castoris to imperial ideology was the dynastic succession. When Tiberius
rededicated the temple in his name and that of his brother Drusus, he claimed for the
imperial family a monument that had had an important political function during
the late Republic and a foundation myth that linked it to one of the great military
victories in the history of the early Republic. Whether the temples late Republican
political function continued into the Principate we cannot know for certain. The
temples foundation myth, on the other hand, was frequently retold and adapted
throughout Roman history and evolved into a national myth. Furthermore, the
traditional mythology of the Dioscuri as demigods, who were elevated to divine statue
as a result of grand achievements and who shared their immortality, made them
suitable divine counterparts to Augustus successors, first C. and L. Caesar, who led
cavalry divisions in the transuectio equitum as principes iuuentutis, and later Tiberius
and Drusus. The temples foundation myth was further adapted to the circumstances
surrounding Drusus death. Tiberius journey to retrieve Drusus body evinced a
fraternal devotion that made their identification with the Dioscuri even more explicit.
The Dioscuris later function as protective deities who oversaw the preservation of the
Concordia Augusta made them especially concerned with the succession, since it was
at the very moment of succession that the harmony of the imperial family was most at
issue. The appearance at Caesars funeral demonstrated to an imperial audience that
the Dioscuri had such concerns even at the very beginning of the dynasty, when the
uncertainty that gripped the city following Caesars death ultimately yielded to the
emergence of the new princeps. Thus, a national myth, comprised of memories
associated with the Aedes Castoris, was transformed into Augustan ideology. The fact
of this transformation shows again how Augustus and his successors exploited traditional elements of the Republic in this case the memories evoked by a Republican
monument to help establish and consolidate their new form of government.
Mount Holyoke College

GEOFFREY S. SUMI
gsumi@mtholyoke.edu

93 K. Scott, Drusus, nicknamed Castor , CPh 25 (1930), 15561 and The Dioscuri and the
imperial cult, CPh 25 (1930), 37980.
94 Tac. Ann. 2.43.6: sed fratres egregie concordes et proximorum certaminibus inconcussi.
Contrast the relationship of Germanicus and the younger Drusus with the relationship of two
other brothers, Nero and Drusus Caesar, the twin sons of Germanicus, whose rivalry only exacerbated the tension within the imperial household (Ann. 4.59.360).

Classical Quarterly 59.1 187195 (2009) Printed in Great Britain

doi:10.1017/S00098388090000147

187
MUTATI
ARTUS
IRENE
PEIRANO

MUTATI ARTUS: SCYLLA, PHILOMELA AND THE

END OF SILENUS SONG IN VIRGIL ECLOGUE 6
Quid loquar aut Scyllam Nisi, quam fama secuta est
candida succinctam latrantibus inguina monstris
Dulichias uexasse rates et gurgite in alto
a! timidos nautas canibus lacerasse marinis;

(Virgil, Ecl. 6.747)

The last section of Silenus song in Eclogue 6 contains a notorious crux.1 Virgil is here
describing the Homeric sea monster Scylla and her white loins girded by barking
monsters. Although his description identifies this Scylla with the Homeric creature
described in Odyssey 12.857, Virgil prominently refers to this character as the
daughter of Nisus (Scyllam Nisi, 74). In doing so, he appears to confuse the Homeric
Scylla with Scylla, the daughter of the Megarian king Nisus. The Megarian Scylla is
known for her betrayal of her father whose magic lock she cut and handed over to
Minos, who at that time was besieging her city, and whose metamorphosis into the
bird ciris is narrated by Virgil in Georgics 1, by Ovid in Metamorphoses 8, and in the
pseudo-Virgilian Ciris.2 This character is thus distinguished from the Homeric Scylla,
daughter of Phorcys and Cratais, who, according to a version found not in Homer but
rather in later sources, was transformed into a monster by Circe in her anger at being
rejected by Scyllas suitor, Glaucus.3
The conflated version of Scylla from Eclogue 6, which explains the Homeric sea
monster as the product of the transformation of Nisus daughter, becomes popular in
the Augustan period.4 It is surely significant that one and the same poet can
* I am grateful to Nicholas Horsfall, Chris Kraus, Richard Tarrant, Richard Thomas and to
CQs anonymous referee for their many insightful comments, criticisms and suggestions.
1 R.F. Thomas, Voice, poetics, and Virgils sixth eclogue, in J. Jasanoff, H.C. Melchert and
L. Oliver (edd.), Mr Curad: Studies in Honor of Calvert Watkins (Innsbruck, 1998), 66976,
reprinted in R.F. Thomas, Reading Virgil and his Texts (Ann Arbor, 1999), 28896; J.J. OHara,
Callimachean influence on Vergilian etymological wordplay, CJ 96 (2001), 369400, at 3924,
and J.J. OHara, True Names. Vergil and the Alexandrian Tradition of Etymological Wordplay
(Ann Arbor, Michigan, 1996), 1445, are the latest treatments of the passage. On the references
to Scylla in the Aeneid, see also M. Paschalis, Virgils Aeneid. Semantic Relations and Proper
Names (Oxford, 1997), 147 and 186.
2 See also Aesch. Cho. 61322; Apollod. Bibl. 3.15.8; Hyg. Fab. 198.
3 Homer (Od. 12.124) has Cratais as Scyllas mother. For different versions of her lineage, see
Ap. Rhod. Argon. 4.828 with Schol. in Ap. Rh. ad 4.82531 g. The story of Scyllas transformation
at the hands of Circe is found in Ov. Met. 13.73014.222 and Hyg. Fab. 199, and seems to have
been told by the third-century poetess Hedyle (SH 456; cf. Ath. 7.297b). On Glaucus as a subject
matter of poetry, see A.S. Hollis, Fragments of Roman Poetry c. 60 BCAD 20 (Oxford, 2007),
1523. For other references to the Homeric Scylla, see Eur. Med. 13434 and 13589; Catull.
64.156 Scylla rapax; Lucr. DRN 5.8923: aut rabidis canibus succinctas semimarinis / corporibus
Scyllas.
4 See Prop. 4.4.3940; Ov. Am. 3.12.212, on which see J.C. McKeown, Ovid Amores 3.12,
PLLS 2 (1979), 16377; Ars Am. 1.3312; Rem. Am. 737; Her. 12.1235; Fast. 4.500, on which see
S. Hinds, Cave canem: Ovid, Fasti 4.500, LCM 9 (1984), 79. See further, S. Timpanaro, De cirri,
tonsillis, tolibus, tonsis et de quibusdam aliis rebus, MD 26 (1991), 10373, at 11718, reprinted
in S. Timpanaro, Nuovi contributi di filologia e storia della lingua latina (Bologna, 1994), 87164,
at 1012; R. Degli Innocenti Pierini, Due note sul mito di Scilla (in Ovidio e nella Ciris), AR 40
(1995), 727, and C. Tsitsiou-Chelidoni, Nomen omen: Scyllas eloquent name and Ovids reply
(Met. 8, 6151), MD 50 (2003), 195203.

188

IRENE PEIRANO

indiscriminately refer to the pure and to the contaminated Scylla at different points of
his career, as Ovid and Propertius do.5 Thus, it is not by mistake that the Augustan
poets refer to the Homeric monster as the daughter of Nisus, since they are equally
capable of producing the conventional version of the story of the Megarian heroine,
as is Virgil in Georgics 1.6 It seems pretty clear that the conflated Scylla found in
Propertius and Ovid is a Virgilian affectation.7 However, this does not exclude the
possibility that Virgil found the conflated version in a source which has not survived.
Evidence for the existence of one or more such sources is scant, but nevertheless
compelling. Ovid, for example, emphatically declares that Scylla, by whom he means
the conflated version of her legend that we find in Eclogue 6, owes her existence to the
poets (Ov. Am. 3.12.212): per nos Scylla patri caros furata capillos / pube premit
rabidos inguinibusque canes. The author of the Ciris complains that many poets
misguidedly told the story of the Megarian Scyllas transformation into the sea
monster.8 Tibullus refers to the story of Nisus coma in a passage illustrating the
power of poetry to confer immortality.9 These references hint at the existence of a
distinguished poetic source for this hybrid Scylla, one potentially pre-Virgilian.10
An unplaced fragment of Callimachus Hecale may offer evidence for the existence
of a conflated version of Scylla already in the Hellenistic period:
(Hecale fr. 90 Hollis / 288 Pf.)
Scylla shameful woman with a name that is not a lie cut the purple lock

The fragment refers to Scyllas cutting of Nisus lock, thus identifying without a
doubt the named character with the Megarian heroine. The phrase
, with a name that is not a lie, clearly signposts an etymological wordplay on
Scyllas name, even suggesting the word etymology.11 So how is Scyllas name true to
her character?
5 Propertius, for example, tells of Scylla, daughter of Nisus, at 3.19.218, and mentions the
Homeric sea monster at 2.26.534, crede mihi, nobis mitescet Scylla nec umquam / alternante
uorans uasta Charybdis aqua, and 3.12.28, Scyllaque et alternas scissa Charybdis aquas. On the
story of the sea monster Scylla in Ovid Met. 13.73014.74, see S.K. Myers, Ovids Causes.
Cosmogony and Aetiology in the Metamorphoses (Michigan, 1994), 98104, and N. Hopkinson,
Ovid Metamorphoses 13 (Cambridge, 2000), 413; cf. also Met. 7.625, on which see S. Hinds,
Medea in Ovid: scenes from the life of an intertextual heroine, MD 30 (1993), 947, at 1121 (on
the relation between this passage and Her. 12.123), and A. Michalopoulos, Ancient Etymologies
in Ovids Metamorphoses: a Commented Lexicon (Leeds, 2001), 1578. Met. 8.1151 is devoted
to the tale of Nisus daughter.
6 Discussion in Thomas (n. 1).
7 See, in particular, Prop. 4.4.3940: quid mirum in patrios Scyllam saeuisse capillos, / candidaque
in saeuos inguina uersa canis? Cf. candida inguina in Ecl. 6.75. See G.O. Hutchinson, Propertius.
Elegies Book IV (Cambridge, 2006), 126 on 4.3940.
8 Ciris 547: complures illam magni, Messalla, poetae / (nam uerum fateamur: amat Polyhymnia
uerum) / longe alia perhibent mutatam membra figura / Scyllaeum monstro saxum infestasse uoraci.
9 Tib. 1.4.636: carmine purpurea est Nisi coma; carmina ni sint, / ex umero Pelopis non nituisset
ebur. Quem referent Musae, uiuet, dum robora tellus, / dum caelum stellas, dum uehet amnis aquas.
The story of Pelops and Tantalus is referred to as a hackneyed topic in G. 3.67: cui non dictus
Hylas puer . umeroque Pelops insignis eburno.
10 Thus McKeown (n. 4), 169, Ovids phrase per nos (through us poets) (21) is therefore a sly
hit at his predecessors.
11 with a name that is not a lie, i.e. with a true name (etumon): OHara (n. 1 [1996]), 77.
R. Pfeiffer on Callimachus fr. 288 points to the Platonic derivation of this phrase: Ap. 34e; Cra.
38c; Plt. 281b. Cf. Virgil, G. 3.2801: hic demum, hippomanes uero quod nomine dicunt / pastores,

MUTATI ARTUS

189

By far the most obvious etymology of the name Scylla is one which arguably
connects the Megarian Scylla to the Homeric monster. As Pfeiffer argued in his note
to the fragment, Callimachus is here suggesting as an etymology of the name Scylla
, a Greek word for young dog.12 The etymology of Scylla from
the noun
skylax, if evoked, must bring about a connection with the Homeric sea monster, as
some commentators have reluctantly acknowledged.13 For the etymological word play
which derives Scylla from skylax was first made by Homer himself in the Odyssey
), and
when he compared Scyllas horrible voice to that of a newborn puppy (
appears in several late grammatical sources.14 In addition, it is surely significant that
Callimachus should choose to play on the etymology of Scyllas name at all, as the
other attested sources which treat the story of the Megarian Scylla focus instead on
the origin of the word ciris, the name of the bird into which she is transformed, quite
possibly to avoid confusion with the Homeric Scylla whose well-established canine
etymology acts as a reminder of her bitch-like features.15 As first noted by Shechter,
, which refers to
the noun Scylla could be seen to be connected with the verb
hair shaving and hair dishevelling, but no ancient source mentions this derivation.16
On the other hand, other attested etymological explanations clearly apply to the
is to trouble
Homeric monster. To start with, the basic meaning of the verb
or annoy, a term which well describes the Homeric Scyllas habit of harassing the
sailors who pass near her.17 Virgil himself alludes to this etymological connection
between Scylla and
by using the verb uexo in line 76 (Dulichias uexasse rates)
of the sea monsters destructive action at the expense of Ulixes ships.18 Also relevant
(
) meaning to strip off,
is a possible derivation from the verb
despoil, glossed by Roman authors with the epithet rapax by which the sea monster
is sometimes described.19
lentum distillat ab inguine uirus pointing to the etymology of hippomanes from hippos and mania
(madness of mares). See S. Shechter, The aition and Virgils Georgics, TAPA 105 (1975), 34791,
at 3634. On etymological signposts, see further Michalopoulos (n. 5), 45.
12 Pfeiffer on Callimachus fr. 288: Call. formula fere Platonica usus dixisse uidetur Scyllam
uero nomine canem esse, i.e. impudentem.
13 Timpanaro (n. 4), 116, n. 29 = 102, n. 29: a malincuore lidentificazione andr accettata;
Tsitsiou-Chelidoni (n. 4), 2023 is more tentative.
14 Od. 12.857:
/
/
. This etymology was recognized and
discussed by grammarians: Schol. Plat. Epist. 7, 345E

(
); Orion in Etym. Magn.
s.v.
:
; Oron in Etym. Magn. s.v.
:
. See OHara (n. 1 [2001]), 3935 for discussion.
15 The Megarian Scylla is turned into the bird known in Greek as ciris whose name reminds
generations to come of her cutting (
) of Nisus purple lock, as Ovid makes explicit at Met.
8.1501: in auem mutata uocatur / Ciris et a tonso est hoc nomen adepta capillo; cf. Ciris 488: esset
ut in terris facti de nomine ciris. In the passage devoted to Scylla in G. 1, Virgil suggests the name
of the bird and its etymology by repetition of the verb seco: secat aethera pennis, 406; secat
aethera pennis, 409.
16 Shechter (n. 11), 359 cites Nic. Alex. 410:
; OHara (n. 1 [2001]),
393 cites Mel. AP 5.175.5:
. With this
etymology signalled by the verb
, the reader would be invited to gloss Scylla as she who
shaves, thereby connecting her name with her infamous cutting of her fathers lock.
17 Beda, Gramm. Lat. 7.289.9: Scylla habet nomen a spoliando siue uexando nautas: spolio enim
et uexo Latine, Graece dicitur scyllo.
18 OHara (n. 1 [2001]), 393, n. 79, notes that the verb caused concerns among the ancient
commentators: see Servius ad Ecl. 6.76; Gell. 2.6.2; Macrob. 6.7.4.
19 A point made by A. Michalopoulos, Some etymologies of proper names in Catullus,

190

IRENE PEIRANO

Commentators have traditionally been reluctant to see a reference to the Homeric

monster behind the etymologizing of the name Scylla, and to acknowledge what
would appear to be a case of contaminatio on the part of Callimachus. Since the
image of a dog was used figuratively of a shameless woman, it is argued that
Callimachus is simply pointing to a connection between Scyllas shameless nature, as
, and her name.20 Thus, while accepting the
described by the participle
standard etymological wordplay connecting Scylla to skylax, proponents of this
theory argue that the latter is to be understood as a synonym for kun and need not
trigger an association with the other Scylla, the Homeric monster with the voice of a
yelping skylax. In his commentary, for example, Hollis acknowledges that Callimachus reference to Scyllas true name must point to a connection between the
characters name and the bitch-like lasciviousness which leads her to cut her fathers
lock and betray her fatherland. However, he is loath to accept that Callimachus may
be implicitly referring to Scyllas transformation into a sea monster with the voice of a
, commenting there need be no suspicion that Call. has conflated the
yelping
two heroines. In addition, Callimachus, it is argued, may have told the correct story
of the Megarian Scyllas transformation into the bird ciris in the Aetia, if we are to
judge from the scant remains of an unplaced fragment (fr. 113 Pf. / 63 Massimilla), in
which Pfeiffer argued the name ciris can be restored, together with the Greek noun
oinos.21
It all depends, however, how one approaches the issue of mythological lore in the
original Callimachean source. For sure, it is not desirable to see the doctus poeta,
Callimachus, making such a blatant mistake. It is, however, equally undesirable to
exclude a readily available connection with the Homeric Scylla and her puppy-like
voice which the clever and learned Callimachean implied reader could hardly be
expected to miss. Furthermore, there is good evidence that play on mythological
homonyms was a standard feature of the learned poetry of the Hellenistic period.
Such play frequently included conflation and simultaneous allusion to multiple
etymologies and it is against this background that a contaminated version of the story
of Scylla may be best understood.22 By evoking the standard etymology of the noun
PLLS 9 (1996), 7581, at 767, and Michalopoulos (n. 5), 157. For Scylla as rapax see Catull.
64.156; Culex 331; Ov. Met. 7.65 and Her. 12.123. Is Ovid alluding to this derivation in Met. 8.
857 (fatali nata parentem / crine suum spoliat, praedaque potita nefanda / fert secum spolium
sceleris)? Line 87, which is omitted by some manuscripts, is highly suspect (it is considered an
interpolation by the latest editor of the Metamorphoses, R.J. Tarrant), so the argument is inconclusive: there is some discussion in Tsitsiou-Chelidoni (n. 4), 197.
20 For the history of the concept of a shameless woman as a bitch, see the note in West on
Hes. Op. 67 and Fraenkel on Aesch. Ag. 1228. Cf. Hom. Il. 1.225; 3.180 and 6.344. Scylla is
described as
in Aesch. Cho. 621. Garvie on Cho. 61921 points to the fact that the
word is a hapax legomenon and suggests that Aeschylus plays upon the etymology of Skulla,
connecting it, as did Homer (Od. 12.85ff.) with skylax, a young dog.
21 G. Massimilla, Callimachus Aetia 12 (Pisa, 1996), 3745 on fr. 63, places it tentatively in
the first book. There are a number of reasons to suspect that the ciris version of the story found
in Virgil, Ovid and the Ciris, is a Hellenistic creation, perhaps going back to the Aetia fragment
tentatively restored by Pfeiffer, or to some other source. Critics often point to the heavily artistic
and neoteric character of the Scylla passage in G. 1. Cf. R.F. Thomas, Virgil Georgics 12
(Cambridge, 1988), 136 on 1.4049: such concinnity and striving for visual effect is a mark of
neoteric, rather than Virgilian, poetry.
22 Discussion in N.M. Horsfall, Virgil, Parthenius and the art of mythological reference,
Vergilius 37 (1991), 316, at 34. In general, on etymological games in the Hellenistic period, see
OHara (n.1 [1996]), 2142. For a specific example, cf. the play between Argos /Argo in Catull.
64.49, and see R.F. Thomas, Catullus and the polemics of poetic reference, AJP 103 (1982),
14464, at 14852.

MUTATI ARTUS

191

Scylla from skylax which is first found in Homer, Callimachus, I suggest, is here
making a characteristically learned allusion to a conflated version of her story in
which she is transformed into a barking sea monster as a punishment for her lasciviousness. The truthfulness of her name does not simply suggest her lascivious nature,
but rather points forward to the Homeric monster and prophesizes the Megarian
Scyllas subsequent transformation into the sea creature known from the Odyssey.
Evidence for the existence of a conflated version of Scylla to which Callimachus
might be alluding gains further support from careful analysis of the sources. It has not
thus far adequately been noticed how the conflation, much criticized by the poet of
the Ciris, might actually serve a specific aetiological purpose. For there is at least one
important point of contact between Scylla, the daughter of Nisus, and her Homeric
counterpart. According to a scholium to Dionysius Perigetes, Parthenius of Nicaea
told the story of how Minos, upon learning of Scyllas betrayal, tied her to his ships
) along in the sea, whence the sea is called
rudder and left her to be dragged (
Saronic .23 Now, according to Eustathius, in Parthenius Scylla was changed into a
bird.24 Yet, as the author of the Ciris is quick to notice, the natural step for Scylla,
upon being dragged along the sea by Minos, would have been to change into a sea
creature rather than a bird:
Sed tamen aeternam squamis uestire puellam
Infidosque inter teneram committere pisces
Non statuit (nimium est auidum pecus Amphitrites)

(Ciris 4846)

A scholium to Euripides Hippolytus relates the same aetiology of the Saronic gulf
found in Parthenius but explains that after falling into the sea, Scylla, daughter of
Nisus, became a beast and did not change her own nature at all (Sch. in Eur. Hipp.
1200:
). It is quite tempting to see in the word
a reference not to
the elegant ciris, but rather to the monstrous Homeric Scylla.25 Parthenius and his
anonymous Roman follower seem to be correcting an alternative version in which
Scylla, daughter of Nisus, was turned into the Homeric sea monster in the process of
being dragged across the sea.26
Furthermore, Virgils description of the Homeric Scylla in Aeneid 3 would seem to
support the existence of a version in which Nisus daughter is turned into the sea
monster by being dragged. In describing Scyllas destructive action against the ships
at 3.4245, Virgil uses the verb traho:
At Scyllam caecis cohibet spelunca latebris
Ora exsertantem et nauis in saxa trahentem.

Here, Scyllas drawing in of the ships makes up for her being dragged in the sea by
Minos (cf. Propertius 3.19.26, pendet Cretaea tracta puella rate; Ciris 390, per mare

23

Lightfoot fr. 24 a. Text and translation of the fragment from her edition.
Lightfoot fr. 24 b,
.
25 Cf. Eur. Med. 13423:
; Sen. Med.
4078: quae ferarum immanitas, quae Scylla, quae Charybdis; Catull. 60.5 nimis fero corde of the
lover to whom Scylla is mother.
26 R.O.A.M. Lyne, Ciris. A Poem attributed to Virgil (Cambridge, 1978), 299 on 484ff. agrees
that the author of the Ciris is dismissing the variant of the story in which Scylla is metamorphosed into a fish, for which see below.
24

192

IRENE PEIRANO

caeruleum trahitur Niseia uirgo). The Megarian Scylla, once dragged in the sea,
executes her revenge by dragging the unfortunate sailors who sail past her.
Finally, Servius ad Aen. 6.286 attests to the existence of a version in which Scylla
was turned into a fish, as distinct from another version in which she was turned into a
bird.27 He is commenting on Virgils use of the plural Scyllae biformes in his
description of the monsters at the gates of Dis:
Bene plurali usus est numero: nam et illa Nisi secundum alios in auem conuersa est, secundum
alios in piscem.

It is relevant to note here that the passage in Odyssey 12 containing the description of
Scylla was a locus conclamatus of Homeric criticism and one that was therefore likely
to invite ingenious interpretative approaches. The evidence from the scholia shows
that ancient commentators were somewhat puzzled by the barking voice of the
Homeric Scylla and some went as far as deleting the lines in Odyssey 12 containing
the description of Scylla.28 The story of the daughter of Nisus transformation into
the Homeric monster provides an aetiology for the sea creatures canine features, now
explained as reminders of the bitch-like character of the monsters former self. This
phenomenon, whereby a character is punished by being turned into a bird or plant
whose name and features evoke the very sin for which they are paying, is a common
one.29 This conflated version of Scylla would have found support in the tradition of
reading the Homeric Scylla as a symbol of shamelessness as attested by the Homeric
Allegories of Heraclitus, according to whom Scylla represents an allegory of
.30 The allegorical interpretation of the Homeric Scyllas canine features as
symbols of shamelessness thus provides a readily available connection with the lustful
Megarian heroine.
To sum up, I have argued that in the Hecale fragment Callimachus constructs or
alludes to an unusual, learned version of the story in which the Megarian Scylla is
turned into the sea monster known from Odyssey 12 by being dragged across the sea
by Minos. Far from being a mistake, this conflated variant, as I have suggested,
explains the barking voice of the Homeric Scylla as a reminder of the characters
former wickedness and shamelessness.
What remains to be examined is the purpose served by the Callimachean reference
in Eclogue 6, to which I now return. The passage concerning Scylla and Philomela,
which represents the concluding section of Silenus song and as such arguably
occupies a prominent position, gains added emphasis by Virgils refusal to tell the
27 The painter Androcydes of Cyzicus was famous for the meticulous care with which he
depicted the fish surrounding Scylla in a painting representing her story: Ath. 8.341a; Plut.
Quaest. Conv. 4.665d. In art, the Homeric Scylla is often represented with the torso of a beautiful
maiden and the lower body half-made of dogs and half-made of fish, as in the Scylla group from
Tiberius villa in Sperlonga: G.B. Waywell, Scilla nellarte antica, in B. Andreae, C. Parisi
Presicce (edd.), Ulisse. Il Mito e la Memoria (Rome, 1996), 10819.
28 The problem seemed to have been reconciling Scyllas terrifying cry with her puppy-like
voice. See Schol. in Od. 12.86:

29
30

.
Myers (n. 5), 379.
Heraclitus, Alleg. Hom. 70.11,

.
Cf. also Eustathius in Od. 2.8.40, and see N.M. Horsfall, Virgil, Aeneid 3 (Leiden, 2006), 317 on
Aen. 3.427.

MUTATI ARTUS

193

story in full. For the praeteritio (quid loquar), which prominently marks the transition
to the final section of Silenus song, seemingly advertises the fact that the poet is
glossing over a story that has already been told.31 If we assume that Virgil found the
conflated version of the story of Scylla in Callimachus, the rhetorical question could
be interpreted as genuine: what need is there to tell a story that has already been
told?32 On a different level, however, the praeteritio is markedly ironic and selfreferential: by stating that there is no need to tell the relatively obscure tale of Nisus
Scyllas transformation into a sea monster, Virgil draws attention precisely to the fact
that he is using an untraditional and somewhat controversial version of the story.33
From a rhetorical and formal point of view, the praeteritio, with its apparent rejection
of Scylla and Philomela as appropriate subjects of poetry, links the end of Silenus
song to the main narrative frame, and specifically to the recusatio at the beginning of
the poem in which the speaker similarly, albeit for different reasons, declines to sing
about a specific topic (reges et proelia).34 Furthermore, the conflated Scylla of lines
747 with its Callimachean background nicely balances the beginning of the poem
(cum canerem , 25), which opens with an adaptation of the Aetia prologue (fr.
1.234 Pf.). Indeed, Virgil advertises his dependence on a model in several conspicuous ways. First, the reference to fama in line 74 can be regarded as an instance of
the so-called Alexandrian footnote, a device whereby the alluding poet draws
attention to his use of a learned source.35 It may also serve as a distancing device,
giving a clue to the reader that the poet does not necessarily endorse the conflated
version.36
Virgil, however, is undoubtedly doing more than paying homage to his Hellenistic
predecessor. Servius, confronted with the mistake raised by the text, had already
argued that the Virgilian Scylla was either a case of poetic licence (poetarum more) or
a compressed allusion to both stories (hysteron proteron), that of the Homeric Scylla
and that of Nisus Scylla.37 Besides the mistakes, what has attracted the attention of
commentators is the somewhat strained syntax of the passage. The meaning of secuta
31 On praeteritio as a transitional device, see R. Gibson, Ovid Ars Amatoria Book 3
(Cambridge, 2003), 164 on 3.16970, quid de ueste loquar?: the transition to a new subject is
signposted with the self-referential question commonly used in didactic verse in these contexts.
For discussion of this passage, see Z. Stewart, The song of Silenus, HSCP 64 (1959), 179205, at
1956.
32 Cf. Virgil, G. 1. 104 quid dicam ? introducing a markedly Homeric passage describing a
river flood. Thomas (n. 21), 845 on 1.10410, describes the passage as the first of many
extended literary adaptations in the poem.
33 For this ironic use of praeteritio to highlight mythological innovations, see, for example, Ap.
Rhod. Argon. 3.3114:
/
/
/
; where the rhetorical question (
) emphasizes Apollonius use of a
recondite version of Circes story: see R.L. Hunter, Argonautica Book 3 (Cambridge, 1989), 134
on 3.314. Cf. 3.10969 where Jason glosses over the story of Ariadne with a similar rhetorical
question. See also, Eur. Hel. 1423:
.
R.J. Tarrant, Roads not taken: untold stories in Ovids Metamorphoses, MD 54 (2005), 6589, at
667, similarly notices Ovids tendency to display his doctrina by explicitly suppressing as
well-known stories that are just as recondite as the ones he chooses to tell.
34 On the similarity between praeteritio and recusatio, see Tarrant (n. 33), 67.
35 For the term, see D.O. Ross, Backgrounds to Augustan Poetry: Gallus, Elegy and Rome
(Cambridge, 1975), 78.
36 N.M. Horsfall, Virgilio. Lepopea in alambicco (Naples, 1990), 126; cf. Ciris 5491 with Lyne
(n. 26), 125 on 5491.
37 Servius ad Ecl. 6.74: modo ergo Vergilius aut poetarum more miscuit fabulas et nomen
posuit pro nomine, ut diceret Scyllam Nisi pro Phorci aut certe sit hysteronproteron, ut quasi
utriusque fabulae uideatur facere commemorationem.

194

IRENE PEIRANO

est in line 74 has caused the most problems. Is it the fama who has pursued Scylla
(Scylla who has been pursued by the story that she attacked )? Or is sequi being
used intransitively (as fama subsequently developed)?38 The harshness of the syntax,
I argue, is no mistake: the notion of fama pursuing Scylla implicitly refers to the
omitted version in which Nisus pursues (insequitur, G. 1.408) Scylla, as Wendell
Clausen already hinted at in his commentary.39 This learned game whereby a poet
chooses one account of a story but subtly alludes to the omitted version is paralleled
in Ovids treatment of Scyllas legend.40 Furthermore, the verb sequi, which is often
applied to an authors relationship to his sources, draws attention to the process of
artistic imitatio.41
So far I hope to have shown that in Eclogue 6 Virgil is following a version of the
story which conflates the two Scyllas and which retrospectively explains the Homeric
barking sea monster as the product of the transformation of Nisus daughter. This
account, which, as I have argued, he may well have found in Callimachus Hecale, is in
competition with another version in which Scylla, daughter of Nisus, is metamorphosed into a bird, a version which Virgil does not explicitly adopt but implicitly
acknowledges.
Virgils mistakes in this passage, however, do not end with Scylla. It has long been
noticed that Virgil refers to Philomela when mention of her sister Procne would be
expected.42
aut ut mutatos Terei narrauerit artus,
quas illi Philomela dapes, quae dona pararit,
quo cursu deserta petiuerit et quibus ante
infelix sua tecta super uolitauerit alis?

(Virgil, Ecl. 6.7881)

Traditionally, it was Procne, Tereus wife, who punished her husband for the brutal
rape of her sister Philomela by serving him their son Itys for supper.43 Virgil, on the
other hand, makes Philomela, not Procne, the avenger.44 In the Greek tradition,

38 The translation is from R. Coleman, Virgil Eclogues (Cambridge, 1977), 199 on Ecl. 6.745.
J. Conington and H. Nettleship, The Works of Virgil (London, 188398), 823 on Ecl. 6.74,
speaks of the phrase quam fama secuta est as a tame and unmeaning parenthesis.
39 W. Clausen, Virgil Eclogues (Oxford, 1994), 205 on Ecl. 6.74.
40 Degli Innocenti Pierini (n. 4); Hinds (n. 4).
41 Cf. Lucr. DRN 3.3 (of Epicurus) te sequor, o Graiae gentis decus; Plin. Ep. 7.30.5 (of
Demosthenes) Quam sane, cum componerem illos, habui in manibus, non ut aemularer improbum
enim ac paene furiosum , sed tamen imitarer et sequerer. See A. Reiff, Interpretatio, Imitatio,
Aemulatio. Begriff und Vorstellung literarischer Abhngigkeit bei den Rmern (Inaugural Diss.,
Cologne, 1953), 1078.
42 See Thomas (n. 1), 670 = 289. A version of the story in which Philomela, and not Procne, is
Tereus wife is posited by Clausen (n. 39), 206 on 7881: as in the case of the two Scyllas, an
ambiguous tradition permitted V. to conflate the roles of the two sisters. A. Hudson-Williams,
Some passages in Virgils Eclogues, CQ 30 (1980), 12432, at 130, proposes to see Philomela as a
composite figure, symbolizing the two sisters. More recently, in discussing Ecl. 6, S.J. Harrison,
Generic Enrichment in Vergil and Horace (Oxford, 2007), 57, points to a fragment of Parthenius
(fr. 33 Lightfoot) which, however, refers to the standard story of the nightingales mourning for
her dead son.
43 Cf. Ov. Met. 6.41274. D. Hurley, Ovid, Met. 6.640: a dialogue between mother and son,
CQ 47 (1997), 3202, has a useful discussion of the episode and of Ovids adaptation of the tragic
version of Tereus and Procnes story.
44 In line 79 it is Philomela who is serving Itys remains. In Ovid it is Tereus wife (Procne) who
initiates the feast: Met. 6.6479. There is some disagreement as to the subject of uolitauerit in line
81. The epithet infelix would seem to apply more naturally to Tereus but the change of subject is
harsh. It is thus better to take Philomela as the subject of lines 7881.

MUTATI ARTUS

195

Procne is known as Tereus wife and she is transformed into a nightingale, who is
often represented in the act of mourning her son Itys (or Itylus). Philomela, on the
other hand, is known as Procnes sister and she is transformed into a swallow.45 In the
Roman tradition, the identities of the swallow and nightingale respectively are often
reversed, with Philomela being referred to as the nightingale (Virgil, G. 4.51115) and
Procne as the swallow (Ovid, Fast. 2.8536; Ars Am. 2.383 and Tr. 3.12.9 ). In an effort
as
to explain this curious switch, scholars have invoked a folk etymology of
).46 The folk etymology would certainly support a revised
she who loves song (
version in which the character of Philomela is transformed into a nightingale, a bird
to whom poets often compare themselves, particularly in the Hellenistic period.47 If
so, this folk etymology of the name Philomela might also invite the reader of Eclogue
, which means both limb and song.48
6 to gloss the word artus (78) with the Greek
Now, in addition to the unexpected role assigned to Philomela, Virgil has significantly
altered the narrative order of events: the final stage of the story, namely Tereus
transformation (mutatos artus), is mentioned first, followed by the middle stage, the
ill-omened banquet (dapes). The prominent reference to Tereus transformed limbs
(mutatos artus) which opens the very last section of the song should then be read as a
self-conscious commentary by Virgil on his use of modified versions of myths, as well
as on his appropriation and transformation of poetic models in the poem as a whole.
The passage devoted to Scylla and Philomela represents the concluding section of
Silenus song. The carmen, which opened with the story of the creation of the world
and the change and combination of elements to create shapes (rerum paulatim sumere
formas, 36), appropriately ends with the two stories of Scylla and Philomela, whose
main characters are subject to change and transformation. But, by his bold adoption
of innovative and unusual variants of myths in this passage, Virgil further thematizes
change and innovation not just as a subject for his poetry, but as a poetic strategy and
as a way to approach the tradition. It is appropriate, therefore, that in this final
passage, creation is not an act of God or nature happening in a void (magnum per
inane, 31). Instead, the mutati artus of Scylla and Philomela, coming as they do after
Gallus poetic initiation, focus the attention of the reader on poetrys power to
change, recreate and refashion the mythical past.
Yale University

IRENE PEIRANO
irene.peirano@yale.edu

45 See F. Bmer, P. Ovidius Naso Metamorphosen: Kommentar, Buch VIVII (Heidelberg,

1976), 11519 on Met. 6.412674.
46 T.E. Page, Bucolica, Georgica, Aeneis (London, 1895), 148; Coleman (n. 38), 2001 on 6.78.
Further discussion in W. Frentz, Mythologisches in Vergils Georgica (Meisenheim, 1967), 937.
47 Callim. Aet. fr. 1.16 (restored by Pfeiffer) and Epigr. 34. Cf. Ov. Am. 2.6.710.
48 Artus is sometimes glossed as Greek
: Gramm. Lat. 4.578.20. In addition, similarly to
membrum, which can apply to a physical as well as to a literary entity, poets sometimes ascribe to
their own artus the qualities of their poetry: e.g. Ov. Am. 2.10.23 graciles non sunt sine uiribus
artus; Tib. 2.3.9 nec querer quod sol graciles exureret artus; Prop. 2.22A.212 sed tibi si exilis uideor
tenuatus in artus / falleris. On the ancient practice of employing body parts to refer to
grammatical and rhetorical units, see further G.W. Most, Disiecti membra poetae: the rhetoric of
dismemberment in Neronian poetry, in R. Hexter and D. Selden, Innovations of Antiquity (New
York, 1992), 391419, at 4067.

Classical Quarterly 59.1 196211 (2009) Printed in Great Britain

doi:10.1017/S00098388090000159

196
DIDACTIC POETRY AND
DIDACTIC
PROSE
G.O.
HUTCHINSON

READ THE INSTRUCTIONS: DIDACTIC POETRY

AND DIDACTIC PROSE*
If we survey classical literature as a whole, an obvious and fundamental question is
how the two large categories of poetry and prose relate. The question has not, to
understate, been accorded the centrality it merits. There are signs of a greater interest
recently. The present piece considers what may seem a promising test case: didactic
poetry after the second century B.C., and prose related to it. This poetry typically
presents a direct transposition of prose; we can often compare either the original or
the wider tradition on that subject. It will emerge, however, that even in this test case
poetry and prose are less easily and sharply differentiated than has been thought. The
piece does not seek to claim that there are no differences between poetry and prose.
Nor is it forgotten that differentiation could be quantitative: that poetry and prose
could have more or less of a given feature, rather than one possessing that feature, one
lacking it. The aim is rather to show the difficulty of the question and the complexity
of the relationship. It would be a very welcome outcome if this article prompted
firmer attempts at differentiation, and still more welcome if it enhanced appreciation
of what is here called didactic prose: prose with a subject matter comparable to that
of didactic poems.1
Two common views on the transposition of prose into didactic poetry are: that the
poets display their skill by transforming into elegant verse recalcitrant material from
unassuming prose; and that the poets invest mundane subject matter from narrow
prose treatises with metaphorical and wide-ranging significance. Both these views
contain much truth on the poetry; but both underrate the prose, as if it were a
colourless container for intrinsically base matter. Such a view of technical prose is
rightly beginning to be challenged. As will be seen, works of didactic prose, like works
of didactic poetry, show differing degrees of literary ambition; internally, works of
didactic prose, like didactic poems, commonly contain more and less elevated
passages. The traffic between poetry and prose is two-way: prose can draw on poetry
as well as the reverse; poetry can seek to evoke or appropriate characteristics of
prose.2
* Versions of this piece have been tried out in Cambridge, Manchester and Vercelli. I am
grateful to my listeners, and especially to Professors L. Battezzato, P.R. Hardie, and D.R.
Langslow. CQs referee has made many helpful suggestions on presentation.
1 Interest in poetry and prose: e.g. D.R. Langslow, The language of poetry and the language
of science: the Latin poets and medical Latin , in J.N. Adams and R.G. Mayer (edd.), Aspects
of the Language of Latin Poetry, PBA 93 (Oxford, 1999), 183225; T. Reinhardt, M. Lapidge,
J.N. Adams (edd.), Aspects of the Language of Latin Prose, PBA 129 (Oxford, 2005), including
J.N. Adams, M. Lapidge, T. Reinhardt, Introduction, 136, at 24, and H.M. Hine, Poetic
influence on prose: the case of the Younger Seneca, 21137; earlier e.g. G.O. Hutchinson, Latin
Literature from Seneca to Juvenal: A Critical Study (Oxford, 1993). For the relation of didactic
poetry and prose see R.K. Gibson, Didactic poetry as popular form: a study of imperatival
expressions in Latin didactic verse and prose, in C. Atherton (ed.), Form and Content in Didactic
Poetry (Bari, 1998), 6798; M. Horster and C. Reitz (edd.), Antike Fachschriftsteller: Literarischer
Diskurs und sozialer Kontext, Palingenesia 80 (Stuttgart, 2003); M. Horster and C. Reitz (edd.),
Wissensvermittlung in dichterischer Gestalt, Palingenesia 85 (Stuttgart, 2005).
2 On the last point cf. n. 7 below. For general discussion of didactic poetry see W. Kroll,

DIDACTIC POETRY AND DIDACTIC PROSE

197

A straightforward difference in pragmatics is frequently assumed (prose is for use,

poetry not); but evidence on ancient readers should raise some doubts on its firmness:
readers are as significant as authors expertise and conjectured intentions. Various
prose statements on didactic poetry have been taken as straightforward truths about
authors, without thought for the intellectual context which they presuppose, and what
this implies about readers. Even when Hipparchus presents Aratus as humbly
following Eudoxus, it is apparent that many readers are using Aratus as a source of
facts; Hipparchus own work is a commentary on Aratus, not Eudoxus. His point that
Aratus poetic charm makes his assertions seem more worthy of credit (
, 1.1.7) takes us back
to the Pindaric idea of poetry conferring authority often erroneously (Ol. 1.28-32).3
Even when Seneca tells us that Virgil wanted not to teach farmers but to delight
readers (nec agricolas docere uoluit | sed legentes delectare, Ep. 86.15), his argument
(16) clearly presupposes that many disagree on the negative part of the phrase, and
take Virgil as a worthwhile source. Columella sees Virgil as teacher and authority,
while eagerly admiring his poetic ornamentation (Arb. 26.1.1 placet igitur sicut
Vergilio nobis).4
Lehrgedicht, RE 12.184257; E. Phlmann, Charakteristika des rmischen Lehrgedichts,
ANRW 1.3 (1973), 813901; B. Effe, Dichtung und Lehre. Untersuchungen zur Typologie des
antiken Lehrgedichts (Munich, 1977); A. Schiesaro, P. Mitsis, J. Strauss Clay (edd.), Mega nepios.
Il destinatario nellepos didascalico (Pisa, 1993) = MD 31 (1993); A. Dalzell, The Criticism of
Didactic Poetry: Essays on Lucretius, Virgil, and Ovid (Toronto, Buffalo, London, 1996); P.
Toohey, Epic Lessons: An Introduction to Ancient Didactic Poetry (London, 1996); Atherton (n.
1); D.P. Fowler, The didactic plot, in M. Depew and D. Obbink (edd.), Matrices of Genre:
Authors, Canons and Society (Cambridge, MA, 2000), 20519; K. Volk, The Poetics of Latin
Didactic: Lucretius, Vergil, Ovid, Manilius (Oxford, 2002), and Aetna oder Wie man ein
Lehrgedicht schreibt, in N. Holzberg (ed.), Die Appendix Vergiliana. Pseudepigraphen im
literarischen Kontext, Classica Monacensia 30 (Munich, 2005), 6890; M. Gale (ed.), Latin Epic
and Didactic Poetry: Genre, Tradition and Individuality (Swansea, 2004); P. Kruschwitz and
M. Schumacher, Das vorklassische Lehrgedicht der Rmer (Heidelberg, 2005); the implications of
P.R. Hardie, Political education in Virgils Georgics, SIFC 4th ser. 2 (2004), 83111, could be
extended. On didactic prose see M. Fuhrmann, Das systematische Lehrbuch. Ein Beitrag zur
Geschichte der Wissenschaften in der Antike (Gttingen, 1960); B. Meiner, Die technologische
Fachliteratur der Antike. Struktur, berlieferung und Wirkung technischen Wissens in der Antike
(ca. 400 v. Chr.ca. 500 n. Chr.) (Berlin, 1999); C. Nicolet (ed.), Les Littratures techniques dans
lantiquit romaine. Statut, public et destination, tradition, Entretiens de la Fondation Hardt 42
(Geneva, 1996); W. Kullmann, J. Althoff, M. Asper (edd.), Gattungen wissenschaftlicher Literatur
in der Antike (Tbingen, 1998); D.R. Langslow, Medical Latin in the Roman Empire (Oxford,
2000); S. Goldhill, The Invention of Prose, Greece and Rome New Surveys 32 (Oxford, 2002),
ch. 4; T. Fgen (ed.), Antike Fachtexte (Berlin, 2005); R.G. Mayer, The impractibility of Latin
Kunstprosa , in Reinhardt, Lapidge, Adams (n. 1), 195210; D. Paniagua Aguilar, El
panorama tcnico-cientfico en Roma (siglos III d.C.). Et docere et delectare (Salamanca,
2006).
3 Philodemus, De poematis 5 is interesting for ancient debate on what differentiates poetry and
prose (cf. C. Mangoni, Prosa e poesia nel V libro della Poetica di Filodemo, Cronache Ercolanesi
18 [1988], 12738); important too is Strabo 1, esp. 1.2.36 (C 1518). But the general ancient
discussion cannot be investigated here; still less the modern, though the issues are visible from at
least La vita nuova onwards.
4 Cf. RR 2.9.12, 10.11, 21.2, 3.10.20, 12.5, 15.4, 9.9.4, 9.2.1, where Celsus stands between
Virgil and Hyginus in his literary characteristics. For the seriousness with which later writers take
the Georgics as a technical work, see E. Christmann, Zur antiken Georgica-Rezeption, WJA 8
(1982), 5767; A. Doody, Virgil the farmer? Critiques of the Georgics in Columella and Pliny,
CP 102 (2007), 18097. On the passage of Seneca, see also M.S. Spurr, Agriculture and the
Georgics, G&R 33 (1986), 16487, at 1646. On poetry and prose in Columellas own work,
including his prose reworking of his poetic book, cf. 10 pr. 3, 11.12; see J. Henderson,

198

G.O. HUTCHINSON

Even when Cicero dwells on Aratus and Nicanders ignorance (De Orat. 1.6970),
his comparison with the orator indicates that they are agreed to present their subject
matter effectively; this is confirmed by Balbus use of Ciceros version of Aratus in
ND 2.10414. The alleged proximity of poet and orator (est enim finitimus oratori
poeta |, De Orat. 1.70) is noteworthy for the larger question of poetry and prose.5
Vitruvius places Lucretius, Cicero and Varro on a level as exponents of their
subjects: plures post nostram memoriam nascentes cum Lucretio uidebuntur uelut coram
de rerum natura disputare, de arte uero rhetorica cum Cicerone; multi posterorum cum
Varrone conferent sermonem de lingua Latina (9 pr. 17). Even the Ars Amatoria,
however entertaining, is not self-evidently without use: books have, after all, been
written on personal relationships. Ovid in exile does not point Augustus to generic
rules excluding didactic poetry from practical application. In general, modern critics
of poetry have not been inclined to question the standard assumptions: they have
found aloofness from practicality and from a distasteful type of prose too plausible an
attribute of Hellenistic and post-Hellenistic poetry. But the segregation of poetry and
prose is not so simple.6
The present piece does not scrutinize the basic atoms of poetic and prose language;
nor does it embrace the cosmos of whole works, and survey how the structures of
didactic poetry and prose relate and interact. It merely considers a series of passages
in various areas, with an emphasis on some aspects and features of style. None the
less, these stylistic aspects will be seen to connect with wider questions of the writers
intellectual concerns, and the division of the world which is so important in didactic
poetry and prose. The passages will also expose to view an involved intertextuality in
both poetry and prose. This lowly approach may help to confront us with the
entanglements that specific phenomena provide, and so complicate and challenge our
own divisions of ancient writing.7
A considerable range of poetry will be discussed, in date and type, and a still more
considerable range of prose. As will be seen, many factors affect the works besides
their didactic genre, especially in prose, where any such genre will be less cohesive
than in didactic poetry. The factors include stylistic fashion, intellectual allegiance
and language (Greek or Latin). But all this only strengthens the argument. Neither
the poetry nor the prose can be viewed as forming a body of material which is simply
defined and contained by its genre; their interaction is the more complex.

Columellas living hedge: the Roman gardening book, JRS 92 (2002), 11033 (and The Roman
Book of Gardening [London, 2004]); S. Diederich, Rmische Agrarhandbcher zwischen
Fachwissenschaft, Literatur und Ideologie, Untersuchungen zur antiken Literatur und Geschichte
88 (Berlin and New York, 2007), 22158. For the vertical lines in some prose quotations see
section I below.
5 Cicero is presumably drawing on a tradition for his assertions about Aratus and Nicanders
expertise (cf. Arat. Vit. I pp. 8.259.1 Martin, II pp. 11.1412.3, but not on the Georgica). With
praeclare at De Orat. 1.69 cf. Col. RR 2.9.12.
6 For the pragmatics of the Ars Amatoria cf., among other works, A. Sharrock, Ovid and the
politics of reading, in P.E. Knox (ed.), Oxford Readings in Ovid (Oxford, 2006), 23861, and
R.K. Gibson, S.J. Green, A. Sharrock, The Art of Love: Bimillennial Essays on Ovids Ars
Amatoria and Remedia Amoris (Oxford, 2006).
7 The structure of Latin didactic poems in relation to prose, and its intellectual implications,
are considered in G.O. Hutchinson, Talking Books: Readings in Hellenistic and Roman Books of
Poetry (Oxford, 2008), ch. 10; that piece and this have to be read together to see the full scope of
the argument.

DIDACTIC POETRY AND DIDACTIC PROSE

199

I. ASTROLOGY
Astrology presents a rewarding area (for our purposes). Both prose and verse show a
large span of stylistic pretensions. The many horoscopes on papyrus are so plain as to
raise the question of what we allow to reach the level of prose: cf. e.g. P. Mich. 152
)

(second century A.D.; no. 184 Neugebauer and van Hoesen8) (

(
)
|

|
< >
[ ]
|
|
|
[ ]
|
[ ] ( )
|
(In year 24, on the fifth intercalary day
added to the first of the new year; the eighth hour of the night. The Sun in Virgo;
Saturn in Virgo; Mercury in Virgo; Jupiter in Aries; Venus in Libra; the Moon in
Scorpio). But the list form, marked here in the arrangement of the papyrus lines, will
prove as we proceed to have elaborate affinities with both prose and verse; it is a basic
form of instruction and of classical poetry alike.
Vettius Valens, whose nine-book Anthologies were written in Greek in the second
half of the second century A.D., presents horoscopes with little narrative explications
which are to arrest the reader. Despite the many types of writing in his work, he
, 6.9.7 Pingree)
claims in a peroration that he has not written poetically (
unlike some. These poetic writers are divided into actual poets and those who
produce an attractive performance by the arrangement of words (
).9 Valens just labours hard and is
truthful. The ornamental words of others (
, 5.8.110)
adulterate knowledge and lead people astray. For all Valens posing, writers are here
perceived to vary in style; style and even poetry do not necessarily diminish credit
with readers.10
The most ambitious and elaborate writers are Firmicus Maternus in prose and
Manilius in poetry. Firmicus eight-book Mathesis appeared in or before A.D. 337;
Manilius five books were probably published sequentially (cf. 1.11417), the first
after A.D. 9, the fourth probably in the reign of Tiberius (cf. 4.7636). Firmicus might
seem more committed to his career as a writer than to the inculcation of his beliefs: so
one could infer when the pagan Mathesis, written before Constantines death in May
8

O. Neugebauer and H.B. van Hoesen, Greek Horoscopes (Philadelphia, 1959).

Krolls punctuation is better than Pingrees:
shows that
goes with this clause.
10 Intriguing for the different types of astrological writing is the elaborate prefatory letter in
the horoscope P. Lond. 130.134 (first century A.D.; Neugebauer and van Hoesen no. 81),
reminiscent of a literary preface. Further cf. A. Bouch-Leclerq, LAstrologie grecque (repr.
Brussels, 1993); W. Gundel and H.G. Gundel, Astrologumena. Die astrologische Literatur in der
Antike und ihre Geschichte (Wiesbaden, 1966); T. Barton, Ancient Astrology (London, 1994), esp.
5762; B. Bakhouche, LAstrologie Rome (Leuven, 2002). Aratus and his progeny: e.g.
D.B. Gain, The Aratus Ascribed to Germanicus Caesar: Edited with an Introduction, Translation
and Commentary (London, 1976); G.O. Hutchinson, Hellenistic Poetry (Oxford, 1988), 21436;
P. Bing, Aratus and his audiences, in Schiesaro, Mitsis, Clay (n. 2), 99109; R.L. Hunter,
Written in the stars: poetry and philosophy in the Phainomena of Aratus, Arachnion 1.2 (1995)
(http://www.cisi.unito.it/arachne/num2/hunter.html) ; D.A. Kidd, Aratus, Phaenomena: Edited
with Introduction, Translation and Commentary (Cambridge, 1997); E. Gee, Ovid, Aratus and
Augustus: Astronomy in Ovids Fasti (Cambridge, 2000); C. Fakas, Der hellenistische Hesiod.
Arats Phainomena und die Tradition der antiken Lehrepik, Serta Graeca 11 (Wiesbaden, 2001);
M. Fantuzzi and R.L. Hunter, Tradition and Innovation in Hellenistic Poetry (Cambridge, 2004),
22445; W. Hbner, Die Rezeption der Phainomena Arats in der lateinischen Literatur, in
Horster and Reitz (2005) (n. 1), 13354; M. Semanoff, Undermining authority: pedagogy in
Aratus Phaenomena, in M.A. Harder, R.F. Regtuit, G.C. Wakker (edd.), Beyond the Canon,
Hellenistica Groningana 11 (Leuven, 2006), 30317.
9

200

G.O. HUTCHINSON

337, is followed in the 340s by the ferociously anti-pagan De Errore Profanarum

Religionum, written under the more militantly Christian Constans and Constantius
II. Firmicus well illustrates how didactic prose can use didactic poetry as a source.
One of the texts he follows closely is Manilius fifth book. That brilliant satirical
parade is a neglected highlight of poetry in the early first century A.D.; we cannot
know whether literary appeal helped to motivate Firmicus selection.11
Firmicus adaptations are usually shorter than the passages of Manilius on which
they are based. But his writing has a more leisurely and expansive appearance than
Manilius. One brief example is provided by the indefatigable socialite in Manilius
5.646: instar erit populi totaque habitabit in urbe, | limina peruolitans unumque per
omnia uerbum | mane salutandi portans, communis amicus (he will equal a people, and
will live in the whole city [by calling on everyone], speeding over thresholds, and
bearing over them all a single word of morning greeting, a universal friend). Here the
paradoxes are tightly packed. Less so in Firmicus, who with some mistranslation
offers uariabunt semper domicilia; domus sedesque mutabunt, | et per omnium limina |
matutinis semper salutationibus | peruolabunt | (8.6.2) (they will always be changing
their dwelling; they will alter house and abode, and always fly over the thresholds of
all with morning greetings). A point is made, and then elaborated with synonyms
(domus sedesque mutabunt). omn- no longer relates in a tight unit to unum: it is
reinforced by semper, which itself takes up the preceding semper. The verb, an
independent rhythmical unit, forms a forceful climax, and surpasses the previous two
verbs, strategically placed. The whole effect is less compact and more spacious. (The
vertical lines mark rhythmic closes. The difference between prose and the fixed rhythm
of poetry is less firm and more varied than might appear.)12
But we can see Firmicus approach more fully by looking at one of his relatively
extensive sections. In Manilius the Pleiades rising with the sixth degree of Taurus
produce pleasure-going effeminates (cf. 4.51822):
illis cura sui cultus frontisque decorae
semper erit: tortos in fluctum ponere crines
aut uinclis reuocare comas et uertice denso
11 The most recent edition of Firmicus is P. Monat, Firmicus Maternus. Mathesis (3 vols.,
Paris, 19927) (footnotes will give editions and commentaries where works may be less familiar,
or editions recent). See R. Turcan, Firmicus Maternus. LErreur des religions paennes. Texte
tabli, traduit et comment (Paris, 1982), 718 for Firmicus life, 1519 for the paganism of the
Mathesis (more perhaps too generous is M. Edwards, The beginnings of Christianization, in
N. Lenski [ed.], The Cambridge Companion to the Age of Constantine [Cambridge, 2006], 13758,
at 1412); contrast e.g. Err. 16.4 amputanda sunt haec, | sacratissimi imperatores, | penitus, atque
delenda, | et seuerissimis edictorum uestrorum legibus | corrigenda |, 20.7 profanarum rerum strage
gaudentes | exultate fortius |. Discussion of Firmicus use of Manilius, and discussion of Manilius,
esp. 5: e.g. W. Hbner, Manilius als Astrologe und Dichter, ANRW 2.32.1 (1984), 126320;
F. Fontanella, A proposito di Manilio e Firmico, Prometheus 17 (1991), 7592; D. Liuzzi (ed.),
Manilio fra poesia e scienza. Atti del convegno: Lecce, 1416 maggio 1992 (Galatina, 1993); J.-H.
Abry, Manilius et Julius Firmicus Maternus, deux astrologues sous lEmpire, in N. Blanc and
A. Buisson (edd.), Imago antiquitatis. Religions et iconographie du monde romain. Mlanges offerts
Robert Turcan (Paris, 1999), 3545; C. Salemme, Introduzione agli Astronomica di Manilio2
(Naples, 2000); A. Perutelli, Il disagio del poeta didascalico: sui proemi II e III di Manilio, MD
47 (2001), 6784; Volk (n. 2), ch. 6. Less elaborate are the adaptations, say, of Oppian (whose
Halieutica probably appeared A.D. 1778) by Aelian (c. 170235): cf. e.g. Ael. NA 1.32 with Opp.
Hal. 2.253320, Ael. NA 9.66 with Opp. Hal. 1.55483.
12 Cicero, admittedly making an argument, calls the poet numeris astrictior paulo | than the
orator (De Orat. 1.70). On Latin prose rhythm, see e.g. G.O. Hutchinson, Rhythm, style, and
meaning in Ciceros prose, CQ 45 (1995), 48599.

201

DIDACTIC POETRY AND DIDACTIC PROSE

fingere et appositis caput emutare capillis

pumicibusque cauis horrentia membra polire
atque odisse uirum teretisque optare lacertos.
femineae uestes, nec in usum tegmina plantis
sed speciem, fictique placent ad mollia gressus
semper amare parum est: cupient et amare uideri.

150

156 (5.14653, 156)

They will always be concerned with their own appearance and the beauty of their brow: with
arranging their hair so that it is twisted into a wave, or summoning back their locks with bands
or shaping them in a thick mass, or altering their head by adding hair; with polishing their
shaggy limbs using hollow pumice-stones, with hating their manhood, or wishing for smooth
shoulders. Their clothes are womens; the covering of their feet is for show not use; they like a
gait shaped for softness It is not enough to be always in love; they will want to be seen in love.

In Firmicus:
erunt semper nitidi, polita fronte et accuratis uestibus prompti, | quorum inflexi crines
torquantur semper in bostrychos, | ut frequenter appositis alienis crinibus | fictam et compositam
pulchritudinem | mentiantur, | totius corporis formam | uario pigmentorum genere mollientes. |
hi demptis pilis corpus suum in feminei corporis | imaginem transferent, | quorum uestes et13 ad
muliebris cultus similitudinem | excolantur. | hi molliter ambulantes | uestigia sua cum delicata
quadam moderatione suspendunt.14 | amabunt semper aut se amare simulabunt, | et
paenitebit eos quod uiri nati sint.
(Math. 8.7.2, 3)
They will always be gleaming, equipped with polished brow and meticulous attire. Their curled
hair will always be twisted into curls, so that often by adding the hair of others they will lay false
claim to a feigned and factitious beauty, softening the shape of their whole body with various
types of ingredient. By removing their body-hair they will shift their bodies into the likeness of a
womans body; their clothes too will be adorned so as to be like a womans get-up. Walking
effeminately, they will take slow steps with a sort of mincing restraint. They will always be in
love or pretend to be so, and they will be sorry that they were born as men.

The closeness of Firmicus adaptation is clear (contrast Liber Hermetis p. 53.810

Gundel);15 but the stylistic differences are apparent, and might at first suggest generalizable differences between poetry and prose. The short parallel clauses of Manilius are
turned into more elaborate periods. In Manilius the phrasing is pithy; individual
words are charged with point and arrestingly combined. Firmicus typically expands,
while building up a sustained rhetorical emphasis on falsehood which goes beyond
Manilius: so for Man. 5.149 caput emutare he has, with a typical emphatic doublet, ut
fictam et compositam pulchritudinem | mentiantur |. Man. 5.147 tortos in fluctum
ponere crines expands to quorum inflexi crines torquantur semper in bostrychos | (an
unusual and satirical loan-word to close). Man. 5.151 odisse uirum becomes, again
with stress on falsity, corpus suum in feminei corporis | imaginem transferent |; 152
femineae uestes becomes quorum uestes et ad muliebris cultus similitudinem |
excolantur |. The dense fictique placent ad mollia gressus (153) becomes the vividly
mocking molliter ambulantes | uestigia sua cum delicata quadam moderatione
suspendunt |. The ablative adjective + noun is characteristic, the quadam Ciceronian;

13 et is problematic; cf. Kroll, Skutsch, and Zieglers apparatus. Possibly uestes etiam: etiam is
very frequently postpositive in the work, cf. e.g. 1.9.2, 1.10.11; for its function cf. e.g. 1.7.16 ista
itaque confidentia mentis erectus | etiam corporis sui | curam tuitionemque suscepit |.
14 Perhaps suspendent?
15 W. Gundel, Neue astrologische Texte des Hermes Trismegistos. Funde und Forschungen auf
dem Gebiet der antiken Astronomie und Astrologie, ABAW Phil.-hist. Abt. n.F. 12 (Munich, 1936).

202

G.O. HUTCHINSON

the combination of delicata with the ostensibly laudatory moderatione swells out the
Manilius with lively irony.16
Yet at the close of his description, Firmicus takes up Manilius odisse uirum for an
effectively curt reinforcement of his theme: the words quod uiri nati sint are short, the
content is damning. Manilius himself closes with a neat sententia, where the first half
elegantly prepares for the second. Such a sententia belongs to his period, in verse and
prose alike.
And there lies the complication. Firmicus style bases itself particularly on Cicero,
as one might expect at this period. This is illustrated by the abundance of synonyms,
and the reduction in epigrams. With the cum delicata quadam moderatione one may
compare e.g. Cic. Inv. 2.164 cum animi ampla quadam et splendida propositione | or
Dom. 115 cum proiecta quadam et effrenata cupiditate |; even the final phrase brings to
mind Cael. 6 id numquam tam acerbe feret M. Caelius | ut eum paeniteat non deformem
esse natum |. If we compared Manilius with prose writers from his own time or just
after, the contrast between poetry and prose would be considerably lessened. One may
instance, on related subjects, the concise, dense or epigrammatic phrases at Sen. Rh.
Con. 1 pr.8 (e.g. capillum frangere), 2.1.6 (Arellius Fuscus; e.g. conuulneratum
libidinibus, | incedentem, ut feminis placeat, | femina mollius |), 5.6 (e.g. muliebrem
uestem sumpsit), Sen. Ep. 95.201 (e.g. et oleo et mero uiros prouocant |), 114.34 (e.g.
quam cupierit uideri |). But the validity and effectiveness of Firmicus transformation
should be apparent.
It has been said above that Manilius poetry is particularly ambitious; the range of
astrological poetry should be briefly exemplified. Manilius account of the
tightrope-walker spectacularly defamiliarizes. The lines et caeli meditatus iter uestigia
perdet | paene sua et pendens populum suspendet ab ipso (5.6545) play on cosmic
divisions and, through word-play, on the inversion of space (pendens suspendet).
The specifically poetic enjambement perdet | paene wittily focalizes the scene
through the audiences terror. The fourth book of Manetho, probably a distinct
poem, is not later than the third century A.D. (P. Oxy. 2546, P. Amsterdam inv. no. 56).
It depicts the activity more straightforwardly:
|
|
(2879). The use of line-end and enjambement is still forceful (287), and the phrase
is of Gorgianic flamboyance. In the preceding paragraph types
of people are denoted with a huge string of adjectives, nouns and short nominal
phrases. This is characteristic of the fourth book especially, and reminiscent of
astrological prose treatises; it is quite unlike Manilius. But the words and phrases
|
show an inventiveness unlike prose. So 2812:
; here
with tombs in foreign lands is a
a unique word,
a cosmic paradox. And
typical new compound,
even Manilius had a string of infinitive phrases: listing or accumulative structures, as
we have mentioned, possess complex connotations.17

16 Firmicus may presuppose the plain hairstyle for men in which his period was imitating
Manilius: cf. A.T. Croom, Roman Clothing and Fashion (Stroud, 2000), 66.
17 Manetho 4.2879 are cited by A.E. Housman, M. Manilii Astronomicon liber quintus2
(Cambridge, 1937), 82. On the style of the fourth book, cf. C.A.M. Axtius and Fr. A. Rigler,
(
 . (Cologne, 1832), xxiii. With populum suspendet cf.
Stat. Theb. 3.107 populos suspendere. Firmicus has an asyndetic tricolon in the passage parallel to
Manilius paragraph, 8.17.4, cf. the tetracolon at 8.15.2.

DIDACTIC POETRY AND DIDACTIC PROSE

203

Firmicus also adapts the poetry of Anubion, which has lately been resurrected on
papyrus; this elegiac poem, probably in at least four books, was perhaps written in the
first century A.D., and certainly by the second century. Subject headings in F 5
(horoscopes of types) and F 15 Obbink suggest in parts of the poem internal division
at short intervals by readers or author; the asyndeton in F 2 and F 5 suggests internal
division by the author: a less ambitious poem, then, than Manilius. Anubions elegiacs
are simpler than Manilius or even Manethos hexameters. Anubions mention of the
|
orator at F 5 b 9(10?) Obbink
(end of book) is greatly expanded by Firmicus, who eloquently depicts the
orators eloquent power: talis erit orator | ut in modum fulminum | dictorum eius
sententiae | proferantur, | ut pro arbitrio eius | multitudinis animi aut quiescentes
excitentur | aut incensi facile mitigentur | (6.30.22). He goes on, with anaphora, to the
orators impact on posterity, and to the verbal might of Demosthenes. The force of
language is not only conveyed but exhibited: the reader is to admire not only
Demosthenes but Firmicus. Even in this didactic genre, the prose self-consciously
rhetorical prose and not verse conspicuously rises above its poetic source. We have
seen, then, the stylistic range of astrological poetry and prose; the use that prose can
make of poetry as well as poetry of prose; and the difficulties of generalizing about
the relation between poetic and prose style, or of making the opposition between the
two forms of didactic writing too rigidly hierarchical.18
II. HORSES
Another subject, horses, takes further the complicated relationship of poetry and
prose; the treatment of animals brings in elaborate literary and intellectual issues. We
will look at three Latin passages, after some points on the prose tradition which lies
behind them. Didactic prose on horses begins with Simon in the fifth century. The
first-person intellectualism of his very opening is highly characteristic of fifth-century
<
>
prose:
But when he tells us the
(5)
of a good hoof, he says that the hollow hoof
(sounds the cymbal) more.
This use of language, though unpoetic, is no less vivid and imaginative than poetry.
solido grauiter sonat ungula cornu is Virgils version (G. 3.88); Xenophon, referring
explicitly to Simon, reduces his linguistic boldness to
(Eq. 1.3). Xenophon talks directly at the start of his treatise about
its relation to Simons: intertextuality is no less important in didactic prose than in
poetry, even if differently expressed. Xenophon also introduces us immediately to
, 1.1); his
epistemological questions and to the evidence of a horses mind (
treatise is not only practical but searchingly thoughtful. The division of mind and
body stressed here will be important for the poetry and prose that follow.19
18

F 5 b 10 sounds like an orator, cf. what precedes and the title; 8 might recommend
[
, considered by Obbink;
would then be the only basis for Firmicus
thunderbolt. For Anubion see D. Obbink, 45037. Anoubion, elegiacs, in N. Gonis (ed.), The
Oxyrhynchus Papyri 66 (London, 1999), 57109; D. Obbink, Anubio: carmen astrologicum
elegiacum (Munich and Leipzig, 2006). It is notable that while Firmicus, who came from Sicily,
uses the Greek Anubion, his readers need an allusion to Philips oratorical opponent to be spelled
out (Math. 6.30.22 ut manifestius explicemus, | Demostheni |).
19 On philosophical issues involving animals in the ancient world see recently C. Osborne,
Dumb Beasts and Dead Philosophers: Humanity and the Humane in Ancient Philosophy and Literature (Oxford, 2007). Text of Simon: e.g. K. Widdra, 1  34 (Leipzig, 1964),

204

G.O. HUTCHINSON

The sequence Varro, Virgil, Columella on the appearance of the good horse shows
textual relationships clearly. It will be of special interest to see how the poet affects the
subsequent prose-writer. Varro, whose work appeared from 37 B.C. on, begins his
account of horses with his common metapoetic or metaprosaic appropriation of
the subject matter, which Virgil will use extensively: Lucienus, ego quoque adueniens
aperiam carceres, inquit, et equos emittere incipiam (RR 2.7.1). Lucienus gives a long
string of the physical features in the foal which suggest a good horse to come: oculis
nigris, naribus non angustis (5), etc. There is no particular rhetorical force in the list.
Lucienus comes a little later to indications of mind (not explicitly named): equi boni
futuri signa, si cum gregalibus in pabulo contendit, in currendo aliaue qua re, quo potior
sit; si, cum flumen trauehundum est gregi, in primis progreditur ac non respectat alios
(6). The final phrase, with alios at the end, lightly suggests an admirable attitude.20
Virgils account is affected by the sublime connotations of the horse in poetry; the
Aeneid shows a clear perception of the prominence and magnificence of horses in
the Iliad. Not that the aesthetic impact of the war-horse is a phenomenon confined
to poetry: Xenophon explains how to make it
(Eq. 10.15). And while Virgils didactic epic interacts here with
narrative epic, he also relishes words that would not be expected in poetry: spadices
(81, chestnut horses) and giluo (82, dun). Virgils depiction interweaves features of
the body with pieces of behaviour, and infuses physical description with mental and
moral qualities; the poetic language realizes inexplicitly what is more explicit in
Xenophon. So for Varros pectus latum et plenum (5) Virgil has luxuriatque toris
animosum pectus (G. 3.81). His version of the animals behaviour is further made
heroic, like a soldiers (or a poets, cf. 3.612): primus et ire uiam et fluuios temptare
minacis | audet et ignoto sese committere ponti, | nec uanos horret strepitus (779). He
later extends this into a martial scene of known sublimity (so Aesch. Theb. 3914
[simile for a hero]; cf. in prose Xen. Eq. Mag. 3.1113): tum, si qua sonum procul arma
dedere, | stare loco nescit, micat auribus et tremit artus, | collectumque premens uoluit
sub naribus ignem (835). The young horse is already like a war-horse eager for the
fray. Virgil ends the paragraph with mythical horses, some explicitly assigned to Grai
poetae (90); animal and human have hitherto been discreetly assimilated, but the
metamorphosed Saturn (923) mixes animal and divine stridently. Employment of
myth seems emphatically characteristic of poetry; but this distinction is weakened by
Xen. Cyn. 1.117 (!) or Var. RR 2.5.5 (including Jupiters metamorphosis into a
bull).21
414. The text of the opening (if it is the opening) is uncertain (
L:
C (the only MSS):
<
>
Blass). A possible version is offered
above; neither Widdras nor G. Pierleonis (G. Pierleoni, Xenophontis opuscula [Rome, 1933], 299)
seems possible. On Simon and Xenophon: J. Althoff, Form und Funktion der beiden
hippologischen Schriften Xenophons Hipparchicus und De re equestri (mit einem Blick auf
Simon von Athen), in Fgen (n. 2), 23552; on horsemanship: J.K. Anderson, Ancient Greek
Horsemanship (Berkeley, 1961).
20 For the latest edition of Varro, with abundant notes, see J. Heurgon and C. Guiraud, Varron,
conomie rurale. Texte tabli, traduit et comment, 3 vols. (Paris, 197897). For a substantial
treatment of Varros work, see Diederich (n. 4), 2253, 172209, 297368, 41019; cf. also
J.E. Skydsgaard, Varro the Scholar: Studies in the First Book of Varros De Re Rustica (Copenhagen, 1968).
21 On mythology in the Georgics see W. Frentz, Mythologisches in Vergils Georgica, Beitrge
zur klassischen Philologie 21 (Meisenheim am Glan, 1967); M. Gale, Virgil on the Nature of
Things: The Georgics, Lucretius and the Didactic Tradition (Cambridge, 2000), ch. 4. On the
anthropomorphizing language used of animals in the Georgics, see M. Gale, Man and beast in

DIDACTIC POETRY AND DIDACTIC PROSE

205

Some of Columellas work was published before A.D. 65; his account of the horse
essentially reworks Varros. Xenophons approach may well have affected him; the
influence of Virgils depiction is clear. So Columellas version of the desirable chest is
lato et musculorum toris | numeroso pectore | (RR 6.29.2). He is more interested than
Varro in mental qualities, and separates them explicitly from the body as Xenophon
does (cf. 6.29.5, and, on other animals, 6.1.1, 2.2, 37.4). Unlike Varro, he starts with
the mental, separating it overtly from the physical (corporis uero forma, 6.29.2).
cum uero natus est pullus, | confestim licet indolem aestimare; | si hilaris, si intrepidus, si neque
conspectu nouae rei neque auditu | terretur, si ante gregem procurrit, si lasciuia et alacritate, |
interdum et cursu certaminis22 | aequalis exsuperat, si fossam sine cunctatione transilit, | pontem
flumenque transcendit, | haec erunt honesti animi documenta.
(6.29.1)
When the foal is born, one can swiftly judge its character. If it is cheerful, if it is fearless, if it is
not alarmed by seeing or hearing something new, if it runs ahead of the herd, if it excels the
other foals in playfulness and keenness, and in occasional running contests, if it leaps over
ditches without hesitation and crosses bridges and rivers: these will be proofs of an admirable
spirit.

The bridge and the noises come from Virgil; but more importantly the noble valour of
the animal is now conveyed. The sequence of conditional clauses is not just a list but a
rhetorical accumulation, which culminates in the moral main clause; this is apparent
from the first three conditionals. The physical description follows Varros string of
ablatives, but near the end suddenly produces a climax which confers elevation on the
horse and the account: sitque sic uniuersum corpus conpositum ut sit grande, sublime,
erectum (6.29.3). The adjectives have a metaliterary resonance: the writing is conscious of its own ambition. Columella returns to the mental with mores autem
laudantur qui (6.29.4).23
In this case the poetic passage separates itself stylistically from the two prose
passages; but the two prose passages differ from each other. The difference is caused
by the impact of Latin poetry as well as of Greek prose. Poetrys range of language
enables boundaries to be tacitly transgressed; but the later prose passage takes up the
suggestions of the poetic language within a more directly philosophical framework.
III. DOGS (AND DOLPHINS)
Other creatures will bring us further into the division of human from animal, and into
the wide aims and involved intertextuality of prose. An affective element was
important in the passage of Columella; it actually forms part of Xenophons rhetorical purpose in the Cynegetica. He wishes to inspire young men to hunting (1.18,
13.118). His description of dogs in action at 6.1516 presents strings of adjectives
and participial phrases: not an instructive list, but an atmospheric evocation of the
Lucretius and the Georgics, CQ 41 (1991), 41426, esp. 41718. In G. 3.81 more decadent senses
of luxuriat, at least, may be pointedly cancelled out; in 84 cowardly connotations of tremit are
forcefully excluded. fremit is a simpler verb in the simile at Ov. Met. 3.704. For Virgils adaptation
of Varro cf. E.W. Leach, Georgics 2 and the poem, Arethusa 14 (1981), 3548; R.F. Thomas,
Prose into poetry: tradition and meaning in Vergils Georgics, HSCP 91 (1987), 22960.
22
Perhaps read cursus certamine.
23
Columellas emotive and extravagant account of mares passion shortly before (6.27.37) is
strongly affected by Virgil, of whom he says in quoting neque enim poeta licentius dicit (6.27.4).
Contrast Var. RR 2.7.78 (though a drastic anecdote follows at 2.7.9). On Columella, see
Diederich (n. 4), 5368, 20958, 36895.

206

G.O. HUTCHINSON

dogs enthusiasm and tumultuous activity. So:

24
(6.16, when they are near
the hare, they will make this clear to the hunter, shaking their whole bodies along with
their tails, charging against the hare in warlike fashion, racing alongside each other in
competition, running together in toil, uniting, quickly parting, charging once more).
He repeatedly gives the words to be used by the hunter to dogs and others (6.1720),
not because the hunter needs a script but to capture and transmit the mood:
(6.17). Such use of the language of everyday speech
might seem available to prose only; but poetry too can imitate it from a distance (Eur.
metaphorical). Interest in commuBacch. 977
nication and feeling between man and animals will recur in writing on dogs.25
Xenophons treatment of rearing and naming puppies (7.35) guides later
treatments. Arrian was consul c. A.D. 130; his intertextuality with Xenophon in his
own Cynegetica goes further than, and perhaps draws on, Xenophons with Simon.
Arrian on this subject (Cyn. 20.30.231.2) cites and quotes Xenophon explicitly; but
he deliberately omits some of Xenophons most striking phrases. So at Xen. Cyn. 7.3
puppies should be suckled by their own mothers, not other bitches:

(embraces).
The last emotive and anthropomorphizing phrase is omitted at Arr. Cyn. 30.2.
Xenophons list of suitable names is also omitted (Xen. 7.5, cf. Arr. 31.2); but here a
more intimate intertextuality comes in.
Arrian so identified himself with the adored Xenophon that he adopted his name,
and gave the last name in Xenophons list to a beloved dog of his own (5.6). Stepping
over boundaries of literature as the dog has stepped over boundaries of species, he
climactically names and memorializes the dog along with himself and his work:

<

>

(5.6, Hence I mean not to

hesitate in writing down the dogs very name, so that for the future too has been left
of her, that Xenophon the Athenian had a dog called Rush, most swift and wise and
).26

24 I have changed the standard punctuation

so as to give
more
point and force.
25 Xenophons, and Arrians, treatises are commented on by A.A. Phillips and M.M. Willcock,
Xenophon and Arrian: On Hunting (
) (Warminster, 1999). On hunting see J.K.
Anderson, Hunting in the Ancient World (Berkeley, 1985). The exact text at Xen. Cyn. 6.17 is
uncertain; cf. also 6.19, 20, Arr. Cyn. 18.1. For excited exclamation which evokes hunting cf. Pl.
Rep. 432D23.
26 Some supplement is needed before
; e.g.
would lead better
into
than Castiglionis
. For the last word Hercher proposes the slightly weak
, taken from 5.2, Sykutris
; Professor L. Battezzato attractively suggests
(cf. hilaris of animals?). P.A. Stadter, Xenophon in Arrians Cynegeticus, GRBS 17
(1976), 15767, at 163, n. 15 thinks that Horme was probably male. See that article and id.,
Arrian of Nicomedia (Chapel Hill, 1980), ch. 4 for Arrians treatise and its relation to Xenophon.
For a relationship between authors approaching identity cf. Ennius reincarnation as Homer
(Ann. I.iix; testimonia O. Skutsch, The Annals of Q. Ennius: Edited with Introduction and
Commentary [Oxford, 1985], 1503). On the names of dogs see E. Bcker, De canum nominibus
Graecis (Diss. Knigsberg, 1884).

DIDACTIC POETRY AND DIDACTIC PROSE

207

Horme is made to surpass her species particularly in her command of language:

she has more sounds than any dog Arrian has seen, and communicates with her voice
(5.5). He treats her reaction to mention of a whip as an understanding of the word,
not just the speakers tone (left out until the end). He gives a whole sequence of her
actions here: the reader is to be amused by her vehemence and her quasi-human
behaviour, but also impressed.

(5.5, she goes up to the man who has mentioned it, and crouches and
begs him, and joins her mouth to his as if kissing, and jumps on him and hangs from
his neck and does not let go until the angry man has desisted from his threat).
Epicureans might view human language as an extension of animal sounds (Lucr.
5.105690); but language is commonly seen as a human prerogative (Cic. Inv. 1.5,
etc.). The whole passage is thoughtful as well as touching. Its intensely first-person
quality distinguishes it from didactic poetry, and also from most didactic prose but
not from epigram (e.g. Mart. 1.109.6 hanc [a named dog] tu si queritur loqui putabis).
Some types of poetry can encompass such lowly domestic affection and praise, not
without an element of humour; but didactic prose can choose to follow them. In the
content and the intertextuality of this passage, literariness, emotion and the authors
private world are connected in self-consciously surprising ways.27
In poetry, we may briefly mention Grattius and [Oppian] on puppies; Grattius is a
contemporary of Ovid, [Oppians] four-book poem probably appeared in A.D. 21217.
Grattius begins from Xenophons recommendation of simple food; he adds the
unpoetic maza (308, barley-loaf ), and relishes the word. The theme of deleterious
luxury enables him to rise into the human (humanos non est magis altera sensus sc.
than luxury of taste, 310). He sweeps onward through geography, history and empire,
addressing Cyrus, Greece and Serranus, and culminating in heaven (325). Such
moralizing expansions are common say in Senecas prose Natural Questions (e.g.
1.17.410: history of mirrors and decline). But Grattius epic genre gives
self-conscious point to the ascent, and to the tension between grandeur and relevance:
scilicet exiguis magna sub imagine rebus | prospicies (3267).28
[Oppian], though heeding Xenophon on short names (Cyn. 1.4445), undoes him
on feeding, and takes off into aspiring fantasy. While suckling from other dogs would
indeed be bad, deer, gazelles, a tame lioness, or a she-wolf would be fine (availability is
not discussed). The puppies would become strong and swift,
(1.443). The humanizing
heightens the crossspecies play to close.
A passage not on dogs illustrates strong affectivity, with a rhetorical purpose; it
illustrates too the interest of poetry as well as prose in the borders of human and
animal. Oppians five-book poem on how to fish (probably A.D. 1778) is passionately
27 On epigrams about dogs cf. M. Citroni, M. Valerii Martialis Epigrammaton liber primus.
Introduzione, testo, apparato critico e commento (Florence, 1975), 3314; G.O. Hutchinson,
Propertius: Elegies Book IV (Cambridge, 2006), 112. For Lucr. 5.105690 see G. Campbell,
Lucretius on Creation and Evolution: A Commentary on De Rerum Natura Book Five, Lines
7721104 (Oxford, 2003), 28394.
28 Commentaries on Grattius: P.J. Enk, Gratti Cynegeticon quae supersunt. Cum prolegomenis,
notis criticis, commentario exegetico (2 vols., Zutphen and London, 1918); R. Verdire, Gratti
Cynegeticon libri I quae supersunt (2 vols., Wetteren, 1964); C. Formicola, Il Cynegeticon di
Grattio. Introduzione, testo critico, traduzione e commento (Bologna, 1988). For dogs in Grattius
see J. Henderson, Going to the dogs/Grattius <&> the Augustan subject, PCPS 47 (2001), 122.

208

G.O. HUTCHINSON

opposed to the fishing of dolphins, on the grounds of their mental equality with
humans, and their friendship with humans on equal terms (Hal. 5.41624). Not far
from the close of his poem he depicts dolphin-fishers attacking a she-dolphin and her
two young (5.51988). The mother (dolphin) is likened to a mother (human) at the
sack of a city (5.5535): thus the animal sphere, in reversal of epic norms, is the target
domain or tenor of the imagery, the human the source domain or vehicle. The likeness
is focalized, as often in Oppian, through an ideal spectator:
|
(5.553).29
The mother is given a speech, as she urges one of her children to flee. The device is
used elsewhere by Oppian (as at 2.3057); but it has special force when the attack is on
mankind (
|
, 5.5601). That force
subserves the persuasive ends of the author: the narrators attack is now voiced by the
character. The poem is exploiting direct speech with relation to humans and animals
in an entirely different fashion from Xenophons prose; one could not imagine such a
device even in the Georgics. But its artifice is acknowledged at the end, with no loss of
|
(5.5656). Poetry
pathos:
no less than prose is interested in divisions and difference; it can exploit the possibilities of its language to transgress supposed divisions, not merely in imaginative
exploration, but with intellectual and argumentative point.
IV. MAKE-UP (AND THE GODS)
The questions we have been exploring could be extended into many other areas, and
into other divisions of the world. It can only be mentioned briefly here that on the
gods, say, prose and poetry show a kindred intellectual energy, kindred stylistic
devices and implicit and explicit interaction. So on the causes of lightning, Lucretius
and Cicero alike heap up mocking questions about Jupiter (e.g. Lucr. 6.4045 in mare
qua porro mittit ratione? quid undas | arguit ?, Cic. Div. 2.45 quid enim proficit, cum in
medium mare fulmen iecit?; cf. Sen. NQ 2.42.1). The vigour of the Latin writers, prose
and verse, may be set against the austere eloquence of Epicurus, at least in the
condensed account of the Letter to Pythocles (104,
[makes deductions]). Although lightning is part of the epic conception of Jupiter, we
do not here find Lucretius directly confronting narrative epic (contrast 1.689). His
concern is rather with another kind of carmina, the books of the Etruscan experts
(6.3812). A plurality of divinities wielding thunder (6.38798) suggests Etruscan
belief, not the world of epic (cf. Plin. HN 2.138). It is Cicero who engages most
explicitly with narrative epic: his own account (fr. 10 Courtney) of the portents before
the Catilinarian conspiracy (Div. 1.1722, 2.45). In a complex combination of generic,
personal, political and philosophical elements, he stresses the self-contradiction (urges
me meis uersibus |, 2.45; contra facta tua et contra scripta, 2.46), but he allows an
29 For the simile cf. Hom. Od. 8.52330, with I.J.F. de Jong, A Narratological Commentary on
the Odyssey (Cambridge, 2001), 21617; for the use of
cf. Il. 17.34. [Oppians] remarkable
simile on the confused searching dog and the pregnant girl (Cyn. 1.494505) may be contrasted
with Valerius Flaccus simile comparing Medea about to leave home with a rabid dog (7.1216).
The latest edition of Oppian: F. Fajen, Oppianus: Halieutica. Einfhrung, Text, bersetzung in
deutscher Sprache, Ausfhrliche Kataloge der Meeresfauna (Stuttgart and Leipzig, 1999); see also
N. Hopkinson, Greek Poetry of the Imperial Period: An Anthology (Cambridge, 1994), 18597;
F. Fajen, Oppianus Ciliciensis 19301999, Lustrum 41 (1999), 75104; E. Rebuffat,
. Tecniche di composizione poetica negli Halieutika di Oppiano (Florence, 2001).

DIDACTIC POETRY AND DIDACTIC PROSE

209

impression to linger: the signs of so great an event may in some sense have been true
(2.468).30
In discussing the argument from design, the Stoic Manilius engages explicitly with
Epicurus, but only through his language with Lucretius: ut uoluit credi, qui primus
moenia mundi | seminibus struxit minimis inque illa resoluit (1.4867). While Manilius
here employs compact metaphors of building, Ciceros Stoic speaker had deployed an
elaborate hypothetical simile on the order in a house or the like (ut, si quis, etc. ND
2.15) an argumentative version of a poetic device. This is made to contrast with the
drier abstraction of the Stoic Chrysippus prose (2.16 = SVF III.1012). The treatment
of the gods, then, would further display the importance, and the unpredictability, of
the relation between poetry and prose, and the many factors which can be involved.31
But we may end by returning to the elemental form of the list, which well exhibits
the scope and complication of stylistic relationships between didactic poetry and
prose. Catalogues are an important form in narrative epic; lists bulk large in the least
literary bureaucracy. The relatively unexalted subject of cosmetics illustrates how the
opposition of poetry and prose is less straightforward than we might think. Medical
recipes, under which ancient make-up falls, appear in simple form on papyrus: so
1180 (firstsecond century A.D.), fr. A, col. ii. 246, 324
32
[ ]
|
|

(
) |
(
)
( )

( )( ) |
( )
(Give your face
a good complexion by crushing bitter almonds and mixing them with water; smear
them on Orange face-powder: bark of the root of convolvulus, litharge, helichrysum, schistus: 12 drachmas weight of each; white lead: 36; Minaean red ochre: 8;
vinegar: a sufficiency). But verse is often used for medical recipes. Galen (second
third (?) century A.D.) commends Andromachus and Damocrates (both first century
A.D.) for using verse: verse is easier to remember and harder to pervert, even if it made
30 For the De Divinatione cf., as well as A.S. Pease, M. Tulli Ciceronis De diuinatione (2 vols.,
Urbana, 19203), D. Wardle, Cicero on Divination: De Divinatione, Book 1, Translated with
Introduction and Historical Commentary (Oxford, 2006) (1445, 160 on the discussion in Book 1).
Cf. further M. Beard, Cicero and divination: the formation of a Latin discourse, JRS 76 (1986),
3346. For Etruscans on gods and lightning, cf. S. Weinstock, Libri fulgurales, PBSR 19 (1951),
12253, at 1219; H.M. Hine, An Edition with Commentary of Seneca, Natural Questions, Book
Two (New York, 1981), 38990; J.-P. Jannot, Devins, dieux et dmons. Regards sur la religion de
ltrurie antique (Paris, 1998), 403. On Epicurus Letter to Pythocles, see D. Sedley, Lucretius
and the Transformation of Greek Wisdom (Cambridge, 1998), 11920.
31 The argument in Cicero, based on Cleanthes (SVF I 528), is strictly meant to show how
(true) ideas of the divine come to be formed in human minds; but in Balbus speech it becomes an
argument for the existence of gods (cf. ND 3.1617, 258). On such arguments in the Stoics, see
M. Dragona-Monachou, The Stoic Arguments for the Existence and the Providence of the Gods
(Athens, 1976), esp. 8891; K. Algra, Stoic theology, in B. Inwood (ed.), The Cambridge
Companion to the Stoics (Cambridge, 2003), 15378, at 15665, esp. 1612; J.C. Thom, Cleanthes
Hymn to Zeus: Text, Translation, and Commentary (Tbingen, 2005), 701; D. Sedley, Creationism and its Critics in Antiquity (Berkeley, Los Angeles, London, 2007), ch. 7, esp. 20513. The
comparison with a house comes already in Aristotle (fr. 12 Rose, Cic. ND 2.95); alternatives, and
elaborate detail on the house, appear in Philo, Leg. Alleg. 3.989 (questionably seen since
I. Bywater, Aristotles dialogue on Philosophy , Journal of Philology 7 [1877], 6487, at 824,
as based on Aristotle). For Lucretian language in the lines of Manilius cf. Lucr. 1.57, 5960,
667, 73, 628, 2.708, 5.4534.
32

is given by I. Andorlini, Un ricettario da Tebtynis: parti inedite di

PSI 1180, in I. Andorlini (ed.), Testi medici su papiro. Atti del Seminario di studio (Firenze, 34
giugno 2002) (Florence, 2004), 81118; but the infinitive seems less plausible than the middle
imperative suggested here, not least because of
.

210

G.O. HUTCHINSON

Andromachus less clear than he could have been (Antid. 14.32, 445, 89 Khn;
Damocrates used iambics). This strikingly practical argument for verse is found
already in the second century B.C. (probably) at Ps.-Scymnus 3342: remembering
unbound discourse (
, cf. oratio soluta) is like holding unbound logs.
(The comic trimeter is favoured for clarity.) Medical recipes even have distinctive
metrical tendencies: elegiacs are much favoured. Galen twice quotes entire the
of Andromachus, Neros physician (GDK no.
174-line elegiac recipe-poem
62). Galens Theriaca begins with a scene reminiscent of the introduction to Ciceros
Topica: Galen has found his addressee reading Andromachus, whose work Galen
admires. The pragmatics of prose and verse are not easily separated in this area.33
Ovids poem on make-up derives from this poetic tradition, already seen in
Eudemus elegiacs (SH 412A, secondfirst century B.C.) as well as in Nicanders
hexameters. Here is to be found the origin of his elegiac didactic; the metaphorically
medical Remedia Amoris closes the series. Ovids elaborate prologue approves
make-up, but also limits its importance and value (4350); the opening of Aristotles
Rhetoric is not wholly dissimilar. For all the humour and the wider bearing of the
work, he also takes on the tradition of presenting a list, numbers, and unusual words
(unusual in poetry). Like Andromachus and others he relishes the idea of distant and
diverse origins an element important in the poetry and the ideology of luxury: so
hordea, quae Libyci ratibus misere coloni (Med. 53), cf. Androm. GDK 62.146
; Illyrica quae uenit iris humo (Med. 74), cf. Nic. Ther. 607
. Nicander is here more precise and
evocative, and proceeds to a myth full of irony: Cadmus and Harmonias transformation into fierce snakes (6089).34

33 The reference to
in Gal. Ther. 14.210 Khn indicates the elder Andromachus, and
presumably this work. On Galens approach to Andromachus poetry see P. Luccioni, Raisons de
la prose et du mtre: Galien et la posie didactique dAndromachos lAncien, in N. Palmieri
(ed.), Rationnel et irrationnel dans la mdecine ancienne et mdivale. Aspects historiques, scientifiques et culturels (Saint-tienne, 2003), 5975; S. Vogt, er schrieb in Versen, und er tat
recht daran: Lehrdichtung im Urteil Galens, in Fgen (n. 2), 5178. For lists in epic, cf. now
S. Kyriakidis, Catalogues of Proper Names in Latin Epic Poetry: Lucretius Virgil Ovid
(Newcastle, 2007). 1180 is first fully published by Andorlini (n. 32); for the lines quoted see 94,
10911. For medical recipes, etc. cf. L.C. Youtie, Three medical prescriptions for eye-salves: P.
Mich. Inv. 482, in J. Bingen, G. Cambier, G. Nachtergael (edd.), Le Monde grec. Pense,
littrature, histoire, documents. Hommages Claire Praux (Brussels, 1975), 55563; H. Harrauer
and P.J. Sijpesteijn, Medizinische Rezepte und Verwandtes (Vienna, 1981); M.H. Marganne,
Inventaire analytique des papyrus grecs de mdecine (Geneva, 1981); C. Schulze, Die
pharmazeutische Fachliteratur in der Antike. Eine Einfhrung2 (Gttingen, 2003). For Ps.-Scymnus
see the recent editions of D. Marcotte, Les Gographes grecs i. Introduction gnrale.
Pseudo-Scymnos, Circuit de la Terre (Paris, 2002), and M. Korenjak, Die Welt-Rundreise eines
anonymen griechischen Autors (Pseudo-Skymnos) (Hildesheim, 2003); on the prologue see R.L.
Hunter, The prologue of the Periodos to Nicomedes (Pseudo-Scymnus), in Harder, Regtuit,
Wakker (n. 10), 12340. On elegiacs and medical recipes cf. M.L. West, Greek Metre (Oxford,
1982), 181.
34 For the Theriaca see J.-M. Jacques, Nicandre. uvres ii. Les Thriaques. Fragments
iologiques antrieurs Nicandre. Texte tabli et traduit (Paris, 2002). For the Medicamina see
G. Rosati, Ovidio, I cosmetici delle donne (Venice, 1985), and V. Rimell, Ovids Lovers: Desire,
Difference and the Poetic Imagination (Cambridge, 2006), ch. 2; for the designation see Ars
3.2056, with R.K. Gibson, Ovid: Ars Amatoria Book 3 (Cambridge, 2003), 17980. The title of
the Ars Amatoria points to a prose tradition: cf. Ath. 162 b on Sphodrias (not Sphodrius)
evidently racy
; N. Heinsius, Commentarius in P. Ovidii Nasonis Opera Omnia, ed.
J.F. Fischer (Leipzig, 1758), 259.

DIDACTIC POETRY AND DIDACTIC PROSE

211

What distinguishes Ovids presentation is the robust sense of action. So at Med.

5762 the couplets end lenta iube scabra frangat asella mola contere in haec (solidi
sexta fac assis eat) protinus in cumeris omnia cerne cauis (bid the slow-moving
she-donkey break them on the rough millstone rub those into them (make a sixth
part of a full pound go in) straight away separate them all in hollow baskets).
Andromachus poem has much more the character of an embellished series of nouns.
This difference is a less extreme form of that between Manilius and Manetho. The
poetry in this area ranges widely. The Ovid should be seen, not as bizarrely tedious,
but as a particularly animated version of a widely read type of poetry. This type bears
an elaborate relation to prose, in form and purpose; Ovids poem cannot plausibly be
viewed as simple parody. Play on women, religion, genre is another matter; it
flourishes in the Medicamina, but is typical of Ovids work in other sorts of poem too.
This brief exploration may suggest the complexity of the relations between prose
and poetry. Obviously, didactic prose and didactic poetry differ in expression and
stylistic form: we have seen the exploitation of devices only possible in poetry or in
prose. But we have also seen links and connections between stylistic forms in didactic
prose and verse, and we have seen the enormous range within didactic prose and
within didactic poetry. In the goals they pursue, and in what they offer to the reader,
facile segregation is not possible. Neither type of work has a monopoly on affective or
atmospheric writing, on argument or persuasive ends. Frequent in both are intellectual depth of content and a concern with divisions and their limits. Intertextuality
is fundamental to both prose and verse, but it includes their interest in and impact on
each other. We should not think of didactic prose as the banausic slave, offering lowly
matter which the master can transform into art. It would be apter to think of a
fascinated and flirtatious relationship between the two types of writing: even if one
type notionally has a hierarchical superiority and the two are notionally contrasted,
this does not prevent intricate interactions and many resemblances. Our approach to
ancient didactic writing needs to be broadened, our understanding of the relation
between prose and poetry deepened.
Exeter College, Oxford

G.O. HUTCHINSON
gregory.hutchinson@exeter.ox.ac.uk

Classical Quarterly 59.1 212226 (2009) Printed in Great Britain

doi:10.1017/S00098388090000160

212
SENECA AND FELICIO:
AND PURPOSE
PATRICIAIMAGERY
AND LINDSAY
WATSON

SENECA AND FELICIO:

IMAGERY AND PURPOSE
INTRODUCTION
Senecas twelfth Moral Epistle is a reflection on old age. He begins with a prooemium
in which he narrates a visit to one of his estates near Rome, where an encounter with
various things familiar from his youth, but now grown old, is a potent reminder of his
own incipient decrepitude:
Quocumque me uerti, argumenta senectutis meae uideo. Veneram in suburbanum meum et
querebar de impensis aedificii dilabentis. Ait uilicus mihi non esse neglegentiae suae uitium,
omnia se facere, sed uillam ueterem esse. Haec uilla inter manus meas creuit: quid mihi futurum
est, si tam putria sunt aetatis meae saxa? Iratus illi proximam occasionem stomachandi arripio.
(2) Apparet inquam has platanos neglegi: nullas habent frondes. Quam nodosi sunt et retorridi
rami, quam tristes et squalidi trunci! Hoc non accideret si quis has circumfoderet, si inrigaret.
Iurat per genium meum se omnia facere, in nulla re cessare curam suam, sed illas uetulas esse.
Quod intra nos sit, ego illas posueram, ego illarum primum uideram folium. (3) Conuersus ad
ianuam quis est iste? inquam iste decrepitus et merito ad ostium admotus? Foras enim spectat.
Unde istunc nanctus es? Quid te delectauit alienum mortuum tollere? At ille non cognoscis me?
inquit: ego sum Felicio, cui solebas sigillaria adferre; ego sum Philositi uilici filius, deliciolum
tuum. Perfecte inquam iste delirat: pupulus, etiam delicium meum factus est? Prorsus potest
fieri: dentes illi cum maxime cadunt.
Wherever I turn, I see proofs of my old age. I had come to my estate outside Rome and was
complaining about the cost of repairing a dilapidated building. My manager told me that it
wasnt the fault of negligence on his part, he was doing all he could, but the villa was old. This
villa grew up under my hands: what will become of me, if stones the same age as I am are in such
a state of decay? Angry at him, I seized the first opportunity to let off steam: These plane trees
are clearly being neglected: they dont have any leaves. How knotted and dried up are their
branches, how miserable and rough are their trunks! This wouldnt happen if someone dug
trenches around them and watered them. He swears by my guardian spirit that he is doing
everything, has given them all his care in every respect, but they are old. Between you and me, I
had planted them, I had seen their first leaf. Turning to the door, I said: Who is he, that decrepit
creature? He deserves to have been put in charge of the door, because hes facing out of doors.
Where did you get him from? What possessed you to take up someone elses corpse? But he [the
old fellow] said: Dont you recognise me? I am Felicio: you used to give me little presents at the
Saturnalia; Im the son of your manager Philositus, your little pet. I said: Hes completely
raving: has he become a little boy again, my pet slave even? But certainly it is possible: even now
his teeth are falling out.

The pathetic image of the elderly slave Felicio, now useful for nothing but doorkeeping and as the butt of seemingly callous jokes, has frequently been discussed in
isolation by historians of Roman slavery.1 But the Felicio episode can only be fully
appreciated in its context, that is, as part of an integrated narrative of a visit by
Seneca to his estate. To illustrate the point that the estate and its contents, like Seneca,
1 E.g. T. Wiedemann, Greek and Roman Slavery (London, 1981), 129; H. Parker, Crucially
funny or Tranio on the couch: the seruus callidus and jokes about torture, TAPA 119 (1989),
23346, at 242; K. Bradley, Slavery and Society at Rome (Cambridge, 1994), 139; C. Laes,
Desperately different? Delicia children in the Roman household, in D. Balch and C. Osiek
(edd.), Early Christian Families in Context (Michigan, 2003), 298324, at 303; cf. also A. Setaioli,
Seneca, lo schiavo Felicione e un iscrizione di Velia, Prometheus 24 (1998), 14951.

SENECA AND FELICIO: IMAGERY AND PURPOSE

213

have grown old, three examples are cited, with a progression from the inanimate villa,
to the living but non-human plane trees, to the human slave Felicio.2 In a typically
Senecan use of imagery, all three entities are a metaphor for senescence in general,
and, in particular, the aged condition of Seneca himself.3 At the opening of the
Epistle, Seneca berates his uilicus for the apparently neglected state of his villa and
plantation. But on being told that this is due to old age, his recollection of building
the villa and personally planting the trees brings home to him vividly how old he himself must be.4 Then comes the encounter with the slave. Again, the uilicus is criticized
this time for buying someone elses old and unwanted seruus.5 This slave intervenes,
asking his owner to recognize him as the child who was taken up as a pet (delicium) by
Seneca at the time when he was establishing his estate, Felicio being the son of the
then uilicus, Philositus.6 The recognition is the final element, in addition to the villa
and the plane trees, associated with the early years of the estate: all three contribute to
Senecas disquieting realization that he must be really getting on if things that he can
remember from their earliest days are now old.
In what follows, we will undertake a more thorough-going discussion of the
prooemium to Epistle 12 than has hitherto been attempted even by scholars such as
Henderson7 who have engaged with the literary subtleties of the text. First, we will
subject the passage to a detailed analysis in order to demonstrate how possibly
fictional details are elegantly woven together by Seneca into an artistic entity. Second,
we will examine the passage in the wider context of both the twelfth Epistle and the
Letters as a whole, demonstrating how it anticipates some broader patterns relating to
Senecas protreptic techniques and themes in the work.
LITERARY ARTIFICE IN 12.13
The reading of Epistle 12.13 which is undertaken below attempts to show that
Senecas account of the visit to his farm is to a large extent a literary construct. Since
this might appear to be labouring the obvious, it is vital to note that a significant
number of scholars, commenting particularly on Senecas indifference towards
Felicios plight, take the episode to be real.8 Even those who recognize the literary
character of the work, such as Christine Richardson-Hay in her recent commentary
on Book 1 of the Epistulae Morales, seem to assume that the whole episode is based
2 Cf. R. Coleman, The artful moralist: a study of Senecas epistolary style, CQ 24 (1974),
27689, at 283, n. 4.
3 Coleman (n. 2) notes that it is typical of Seneca to exploit visual details for metaphorical
purposes.
4 Haec uilla inter manus meas creuit: quid mihi futurum est, si tam putria sunt aetatis meae saxa?
ego illas (uetulas arbores) posueram, ego illarum primum uideram folium.
5 Owners could solve the problem of caring for old, useless slaves by selling them off: see T.G.
Parkin, Old Age in the Roman World: A Cultural and Social History (Baltimore / London, 2003),
2201; K. Cokayne, Experiencing Old Age in Ancient Rome (London, 2003), 171, nn. 114 and
219.
6 For adult Romans taking young slaves as delicia, cf. W.J. Slater, Pueri, turba minuta, BICS
21 (1974), 13340, at 1345; H.S. Nielsen, Delicia in Roman Inscriptions and in the Urban
Inscriptions, ARID 19 (1990), 7988, at 813; Laes (n. 1), 3004. Many scholars (e.g. all cited in
n. 1 above except Setaioli; C.D.N. Costa [ed.], Seneca: 17 Letters [Warminster, 1988], ad loc. and
W.C. Summers [ed.], Select Letters of Seneca [London, 1960], ad loc.) have incorrectly assumed
that Felicio was Senecas slave playmate as a child, which would make them contemporaries.
7 J. Henderson, Morals and Villas in Senecas Letters: Places to Dwell (Cambridge, 2004),
1927.
8 E.g. the scholars listed in n. 1 above.

214

PATRICIA AND LINDSAY WATSON

on an incident which really took place and which has been exploited by Seneca as a
vehicle to make a moral point.9 Others, conversely, have assumed that the account is
a fiction, but without attempting to substantiate this position.10 Given this, and
the further fact that discussions of fictionality in the Epistles typically focus on the
broader issue of whether or not the Letters represent a real correspondence,11 rather
than on possibly fictional elements within the individual epistle, it seems worthwhile
to subject sections 13 of Epistle 12 to a close analysis which will show how a number
of ostensibly autobiographical elements are in fact so artfully contrived, so anchored
in literary tradition, that it is legitimate to question whether there is any more than a
soft kernel of fact at the heart of the whole episode.
In arguing this, it is not our intention to suggest that the character of Felicio is a
total fiction, or to deny that Seneca owned and spent time at such an estate as he
describes in the Epistle: land owners regularly visited their properties to inspect the
work of the uilicus for possible shortcomings, as Seneca represents himself doing
here.12 This said, there are many elements in the episode which when examined closely
may be shown to be intensely literary, and it is our contention that only an
appreciation of this will enable us to appreciate fully Senecas artistry in constructing
his narrative.
To begin with an obvious point, Senecas encountering on the farm of three
exempla of old age (the villa, the plane trees, the elderly Felicio) is too convenient for
comfort, inviting the suspicion that at least one of the incidents was invented for the
sake of completing a rhetorical triad.13
Next, let us consider the description of the villa buildings and the plane trees. The
essential reality of this incident has been taken for granted, yet there is something
seriously wrong with the uilicus explanation that their condition is unavoidable due to
their extreme age. The dramatic date of the letter is A.D. 634, when Seneca was in his
sixties.14 But the villa cannot be as old as this, despite the description of the stones as
of my age (aetatis meae), since Seneca says that he built the villa himself (haec uilla
inter manus meas creuit).15 If we postulate that Seneca acquired the property in his
mid to late teens at the earliest,16 this would make the house about fifty years old.
9 C. Richardson-Hay, First Lessons. Book 1 of Senecas Epistulae Morales: A Commentary
(Bern, 2006), 29, 351. Henderson (n. 7) also seems to assume, if we understand him correctly, that
the whole episode is based on a real incident, while subserving a larger philosophical purpose.
10 Cf. e.g. M.T. Griffin, Seneca: A Philosopher in Politics2 (Oxford, 1992), 277 (Felicio and
other) slaves (named) must have existed, but did the scenes described really take place?
11 Cf. M. Wilson, Senecas Epistles to Lucilius: a revaluation, Ramus 16 (1987), 10221, at
119, n. 3 and the bibliography there cited; Richardson-Hay (n. 9), 334 and n. 55 with further
bibliography.
12 Agricultural writers advise landlords to undertake regular inspections of their estates in
person, in order to check on the efficiency of the uilicus: cf. Cato, Agr. 2.24; Columella 1.7.5,
8.20; J. Carlsen, Vilici and Roman Estate Managers until AD 284 (Rome, 1995), 8592.
13 On Senecas liking for grouping exempla in threes, see R.G. Mayer, Roman historical
exempla in Seneca, in Snque et la prose latine (Fondation Hardt Entretiens 36, Geneva, 1991),
14176 at 1556.
14 Seneca was born between 4 and 1 B.C.: cf. M. Griffin, The Elder Seneca and Spain, JRS 62
(1972), 119, at 78, and Griffin (n. 10), 36. This would make him sixty-two at the youngest and
sixty-eight at the oldest in A.D. 634.
15 As NisbetHubbard note in commenting on Horace, Carm. 2.13.23, where it is said that
the planter of the accursed tree on Horaces estate <arborem> sacrilega manu / produxit, manus
suggests the bestowal of personal attention.
16 Cf. Griffin (n. 10), 287, who suggests that the acquisition of the suburbanum goes back at
least to the days when Seneca was listening to (i.e. studying philosophy under) Sotion and

SENECA AND FELICIO: IMAGERY AND PURPOSE

215

Such a villa, built by a wealthy aristocrat, may be assumed to have been solidly constructed, and unless neglected something that the uilicus denies its stones would
hardly be crumbling with age after such a comparatively short period of time.17
Moreover, even supposing that the uilicus was neglectful of his duties, Seneca himself,
if he was really the hands-on landlord that he depicts himself as here and elsewhere
(cf. Ep. 86, QNat. 3.7.1), might be expected to have noticed, and dealt with, any deterioration long before the estate got into the condition portrayed in this passage.18
Similarly, the description of the supposedly superannuated plane trees, with their
parched, knotted bark and complete lack of foliage (nullas habent frondes. Quam
nodosi sunt et retorridi rami, quam tristes et squalid trunci!), strikes a decidedly odd
note. Elsewhere, planes are conspicuous for their abundant summer leaves, which
made them extremely popular as a shade tree, and they are often associated with the
locus amoenus, a setting to which the idea of shade is integral.19 More importantly,
they are noteworthy for their longevity, as demonstrated by their place in stories of
famously long-lived trees, such as the plane in Delphi purportedly grown there by
Agamemnon (Theophr. Hist. pl. 4.13.2).20 Statius patron Melior had a remarkable
specimen, already of a great age (Silu. 2.3), while Martial (9.61) describes one planted
by Julius Caesar and still luxuriant in Martials day. Nowadays, planes are commonly
planted on city streets, not only because of their abundant shade, but because of their
hardiness and resistance to disease and pollution. The plane trees on Senecas estate,
then, should not be in their wretched condition simply through old age. In fact, as
portrayed by Seneca, they resemble modern descriptions of American sycamores (a
species of plane) affected by the fungal disease anthracnose, which, left untreated, can
lead to complete defoliation and a distorted appearance,21 or by the fungus Ceratocystis fimbriata platani, infestation with which results in sparse foliage, small leaves,
and elongated sunken cankers on the trunk and larger branches.22 We must conclude,
Attalus; Sotions floruit was put by Jerome at A.D. 13 (cf. Griffin [n. 10], 37, n. 9). Since Seneca
the Elder died around A.D. 40 (Griffin [n. 10], 33), the property would have legally belonged to
him and have been part of the sons peculium, but it was common for grown sons still under their
fathers potestas to live separately, e.g. Roscius of Ameria lived in the countryside, on some family
farms of which he had the usufruct. See J.F. Gardner, Family and Familia in Roman Law and Life
(Oxford, 1998), 6872; K.R. Bradley, Discovering the Roman Family (New York/Oxford, 1991),
1634.
17 Stone quarried near Rome was soft and buildings which used it did not last more than 80
years, according to Vitruvius (2.7.12, 8.8), who also describes structural techniques for ensuring
that stone buildings would last as long as possible. For the importance of siting villa buildings to
avoid collapse of the structure, cf. Columella 1.5.910. So even if Senecas suburban villa was
made of local stone it should not have fallen into a state of decay after 50 years.
18 Thanks to Miriam Griffin for this point.
19 On plane trees being prized for their shade, cf. Plin. HN 12.612. For shade as a feature of
the locus amoenus, cf. H.-J. Van Dam (ed.), P. Papinius Statius Siluae Book II (Leiden, 1984), 314
(note on Silu. 2.3.3942). See also G. Schnbeck, Der Locus Amoenus von Homer bis Horaz
(Diss. Heidelberg, 1962), 289, 4956.
20 The tree at Delphi is also mentioned by Pliny (HN 16.238) in a list of famous long-lived
trees, which includes some planes. Further examples of long-lived planes from antiquity are cited
by V. Hehn, Kulturpflanzen und Haustiere9 (Berlin, 1911), 296. In more recent times, a plane tree
has stood in front of Kew Palace in Richmond, near London, since 1762 and is still in fine
condition (for a picture see: http://www.rbgkew.org.uk/plants/trees/oldlions.html).
21 S. Nameth and J. Chatfield, Anthracnose leaf blight of shade trees, Ohio State University
Extension Fact Sheet: Plant Pathology (http://ohioline.osu.edu/hyg-fact/3000/3048.html); cf. P.D.
Manion, Tree Disease Concepts (New Jersey, 1981), 1401.
22 G. Moorman, Canker stain of sycamore and London plane, Penn State University, Cooperative Extension, Plant Disease Facts (2006) (http://www.ppath.cas.psu.edu/extension/plant_
disease/cankerst.html).

216

PATRICIA AND LINDSAY WATSON

then, either that the incident of the planes is an invention, just as the famous plane
tree under which Platos Phaedrus was set was less a feature of the literal landscape
than a construction of Platos pen (cf. Cic. De or. 1.28), or else that the plane trees on
Senecas estate really were in a poor condition but that Seneca, in a piece of fiction,
attributes to old age, for purposes of his discussion, symptoms in fact caused by
disease.23
The doubts just thrown up as to the factual nature of the narrative are reinforced
by the extremely literary character of the passage, which is marked by a number of
well-established tropes. As symbols of human old age, the villa and trees are to a
considerable degree personified. The phrase which Seneca uses of the estate, for
instance, inter manus meas creuit, probably means grew up, was reared, under my
care: this is its sense in the only parallel for the phrase which we have found, Apuleius,
Met. 6.22, where Jupiter declares that he was responsible for the rearing of Cupid.24
Moreover, the conceit that the stones are coeval with Seneca (aetatis meae saxa 12.1)
likewise seems to personify the villa: the stone quarried to build the villa was of
course of indeterminable age, but the age of the stones is equated with the age of the
villa, as if the stones, as part of the villa, had their day of birth at the time that the
villa was built and a finite life-span, involving growing old and decaying. In particular,
it is tempting to read the dilapidated building of the suburbanum (aedificii dilabentis
12.1) as a symbol of Senecas own decrepitude, for the image of a collapsing edifice is
often associated figuratively with the idea of old age, and it is an image for which
Seneca shows a particular fondness, as at 30.2, where an old mans body is compared
to a putre aedificium.25 And the symbolic equation of the deteriorating fabric of the
building with Senecas own decaying frame is powerfully reinforced by the epithet
selected to describe the former, putris: this term and its cognates are often used to
characterize the physical disintegration attendant upon old age.26
The plane trees of section 2 are similarly replete with symbolic resonances. The
idea of trees as images of human life was common in Roman thought.27 The pattern
in such symbolic relationships is that the condition of a tree or trees is somehow
mirrored in its human referent. Thus, for example, Pliny records the phenomenon of
two myrtles, one known as the patrician myrtle and the other as the plebeian, which
flourished or withered according to the relative dominance or diminution of the
23 For circumfodere as a remedy for disease, cf. Plin. HN 17.247. It must be admitted that in
modern times the oriental plane (the species of plane known to the ancient world) is said to be
relatively immune to the diseases nowadays affecting the sycamore, but this does not exclude the
possibility that in Senecas day planes suffered from some sort of disease to which they have
become immune over two millennia.
24 Apuleius, Met. 6.22 at tamen modestiae meae memor quodque inter istas meas manus creueris
cuncta perficiam (in the finale to the Cupid and Psyche story, Jupiter says that despite Cupids
insolent behaviour he will arrange his marriage with Psyche because of his clemency and because
he raised him).
25 Cf. Ep. 58.35; De ira 2.28.4; QNat. 6.10.12; Cic. Sen. 72 with Powell; Parkin (n. 5), 341, n.
55. Rhiannon Ash additionally draws our attention to Tac. Hist. 1.27.2 where Otho claims that
emi sibi praedia uetustate suspecta, is perhaps a metaphor for the state of the empire in the hands
of the elderly Galba. Cf. also (in a slightly different context) the image of life as a house at Sen.
Ep. 70.16.
26 Cf. L.C. Watson, A Commentary on Horaces Epodes (Oxford, 2003), on Epod. 8.7 (where
putrida pectora should refer to elderly mothers, not the Parcae), also the figurative use at Prop.
4.5.69 atque animam in tegetes putrem exspirare paternas.
27 Cf. R. G. M. Nisbet, The oak and the axe: symbolism in Seneca, Hercules Oetaeus 1618ff.,
in S.J. Harrison (ed.), Collected Papers on Latin Literature (Oxford, 1995), 20212, who states at
202 the life <of trees> moves in human rhythms.

SENECA AND FELICIO: IMAGERY AND PURPOSE

217

authority of the Senate or people (HN 15.1201). In a comparable effect, the rotten
and barren trees which grow on the quondam site of Troy (Luc. 9.9669) reflect the
decayed glory of the Trojan kingdom.28 Significantly too, Pliny (HN 17.21819)
relates the diseases and death of trees, in terms of both pathology and nomenclature,
to those of humans, particularly those of the slave or lower classes. And in the most
recent account of tree-symbolism, Philip Hardie has demonstrated that the plane tree
of Atedius Melior in Stat. Silu. 2.3 figuratively echoes a number of its owners traits.29
All this suggests that the aged and deteriorated condition of the plane trees has
similarly been contrived as an image of Senecas own decrepitude, a conclusion which
looks the more inviting, since the comparison of trees with people is a conceit for
which Seneca shows an especial fondness in his tragic corpus.30
To turn now to Senecas nuanced use of language, the symbolic equation of the
aged planes with their elderly owner is surely put beyond doubt by his choice of
adjectives to characterize the former: for nodosus, retorridus, tristis, squalidus and
uetulus, while commonly used of trees, are equally at home in the human sphere, not
least in the sphere of human old age. To take squalidus first. Pliny, for example, speaks
of the plebeian myrtle mentioned above as retorrida ac squalida during the era of
senatorial ascendancy (HN 15.121), but persons can equally be squalidi (Plaut. Truc.
933, huncine hominem te amplexari tam horridum ac tam squalidum?; Lucr. 5.956
squalida membra), and the term is often applied, in differing contexts, to the aged:31
cf. Ter. Eun. 2356, hominem / uideo sentum squalidum aegrum, pannis annisque
obsitum; Plin. Ep. 4.9.22, et in procero corpore maesta et squalida senectus; Apul. Met.
6.18, huic squalido seni.32 Again, while tristis is often used, as here, of trees or plants in
a wretched condition,33 or else, in reference to the arbor infelix concept, of trees
which do not yield flowers or fruit,34 the adjective is also self-evidently applicable to
humans. Significantly, it was twice used by Virgil to characterize old age (tristisque
senectus, G. 3.67, Aen. 6.275) passages which Seneca quotes more than once in his
Epistles, suggesting that its association with senectus may be in play subtextually
here.35 The same duality of reference to the human and arboreal realms is likewise
seen in the rare adjective nodosus.36 Used both here and elsewhere of wood that is
knotted or gnarled (cf. Luc. 3.440, nodosa ilex; Plin. HN 16.65, materies [tiliae si
mas est] nodosa; Juv. 8.247, nodosam uitem),37 nodosus can also characterize the
knot-like deformations of the joints caused by gout,38 a disease from which Seneca
28

Nisbet (n. 27), 207.

P. Hardie, Statius Ovidian poetics and the tree of Atedius Melior, in R. Nauta, H.-J. Van
Dam and J. Smolenaars (edd.), Flavian Poetry (Leiden, 2006), 20721.
30 Cf. Nisbet (n. 27).
31 In part perhaps for the sake of the alliteration with senex or senectus? See, in addition to the
examples cited immediately below in the text, Seneca the Elder, Contr. 1.1.19 senex squalidus
barba capilloque; Sen. HF 765 squalidus senex.
32 Based, like the HF passage quoted in the preceding note, on Virgils famous description of
the infernal ferryman, Aen. 6.2989, 304 portitor has horrendus aquas et flumina seruat / terribili
squalore Charon iam senior.
33 Cf. Plin. HN 13.120, 17.33; Sen. Ep. 86.19.
34 Cf. Plin. HN 16.95 non enim omnes (arbores) florent, et sunt tristes quaedam quaeque non
sentiant gaudia annorum, 16.50; Hor. Od. 2.13.11; J. Andr, Arbor felix, arbor infelix, in M.
Renard and R. Schilling (edd.), Hommages Bayet (Brussels, 1964), 3546.
35 Epp. 107.3, 108.24, 29.
36
See J.F. Gaertner, Ovid, Epistulae ex Ponto, Book 1 (Oxford, 2005), on Ov. Ex P. 1.3.234.
37
The reference in the Juvenal passage is to the centurions baton of vine-wood. For nodosus
of wood, cf. also Plin. HN 16.196, 17.176.
38
Hor. Ep. 1.1.31 nodosa cheragra; Prud. Perist. 10.495 nodosa podagra.
29

218

PATRICIA AND LINDSAY WATSON

may possibly have suffered39 and which tended, it seems, to increase in severity with
the onset of old age (Plin. Ep. 1.12.46).40 Since, however, gout was usually caused by
prolonged over-indulgence in food and drink,41 a vice of which Seneca was certainly
not guilty,42 nodosus, if it has a human referent, more probably refers to arthritis,
characteristically an affliction of old age: for writers often speak in medical contexts
of articulorum or neruorum nodi,43 knotty lumps at the joints/sinews, almost
certainly a reference to the disfiguring excrescences around the joints, the so-called
Heberdens nodes, which are characteristic of the degenerative disease osteoarthritis.44
Of particular interest in this connection is Prud. Perist. 10.495, nodosa torquet quos
podagra et artrisis, where nodosa characterizes arthritis as well as gout. Of the five
adjectives used by Seneca to describe the deteriorated condition of his plane trees,
retorridus, which refers to a shrivelled or dried up appearance in trees or plants,45 is,
on the face of it, the least likely to conceal a secondary allusion to Senecas old age.
But the word is occasionally used of living creatures who are getting on in years;46 and
it might suggest the drying up of the vital juices which, according to humoral theory,
was characteristic of old age:47 interestingly, Galen (1.582 Khn) compared old
persons to a plant that, initially moist and soft, gradually dries out. Much more
promising however for our argument is the last of the words applied to the planes,
uetulus. Although uetulus is a technical term for an old tree,48 a passage of Cicero
seems to indicate that the adjective, thus used, is a metaphor derived from the process
of growth and ageing in animals and humans,49 once more suggesting a symbiosis
between the world of man and the world of trees. And that uetulus is meant to call
to mind the physical decline of Seneca himself is strongly suggested by another
consideration. With the exception of the just-noted application to old trees, and a
Catullan-inspired reference to aged wine,50 uetulus is almost invariably used of
elderly living creatures, most particularly old men and women, e.g. Plaut. Merc. 314,
uetulus decrepitus senex; Epid. 666; Cic Att. 13.28.4, Cornificiam uetulam sane et
multarum nuptiarum; Juv. 13.55, si iuuenis uetulo non adsurrexerat. A concealed
allusion to Senecas senescence, in short, seems almost incontestable.
39

Cf. Epp. 53.6, 78.9; Griffin (n. 10), 19.

The sufferer in Pliny in fact contracted the disease, which was hereditary in his family, in his
thirties, but the pain increased greatly as he grew old.
41 See L. and P. Watson, Martial: Select Epigrams (Cambridge, 2003), 320.
42 At least if we believe his protestations of ascetism, for which see Epp. 83.6, 87.15,
108.1516, 123.13. Also relevant: Ep. 108.21 ad fin.; Tac. Ann. 15.45.3, 63.3.
43 E.g. Plin. HN 24.21, 30.110; Marcell. Med. 35.2.
44 For the prevalence of this from middle age onwards in the Roman world, see R. Jackson,
Doctors and Diseases in the Roman Empire (London, 1988), 1767. Pliny (Ep. 8.18.9) describes
graphically the agonies of an elderly sufferer.
45 Cf. Plin. HN 15.121 on the plebeian myrtle, cited above; Sen. De ira 3.15.4 uides illam
arborem breuem retorridam infelicem?; Ep. 86.18 nec magna pars eius quemadmodum in oliuetis
ueteribus arida et retorrida erit.
46 Gell. NA 15.30.1 qui ab alio genere uitae detriti iam et retorridi ad litterarum disciplinas serius
adeunt; Phaedr. 4.2.1617 post aliquot uenit saeculis retorridus (mus) / qui saepe laqueos et
muscipula effugerat.
47 For the loss of sap in the elderly, cf. E. Eyben, Antiquitys view of puberty, Latomus 31
(1972), 67797, at 67982.
48 Lucr. 2.1168 uetulae uitis; Serv. ad Verg. Ecl. 3.11 and G. 4.144.
49 Cic. Fin. 5.39 earum etiam rerum quas terra gignit educatio quaedam et perfectio est non
dissimilis animantium; itaque et uiuere uitem et mori dicimus, arboremque et nouellam et uetulam et
uigere et senescere.
50 Catullus 27.1 minister uetuli puer Falerni initiates the trend.
40

SENECA AND FELICIO: IMAGERY AND PURPOSE

219

If, as suggested above, the first two exempla of old age the decrepit building and
the deteriorated planes are unreal or at least involve an embellishment of reality
with an overlay of literary symbolism and metaphor, then there is no reason why we
should take the Felicio episode as the exact truth either. One element of literary
invention might be the names of both Felicio and his father, both of which sound
suspiciously appropriate in the context. Philositus, lover of grain, is a conveniently
suitable name for a uilicus,51 while the use of the name Felicio, ironic in the context (he
is anything but felix), could be an example of the literary commonplace whereby
proper names are applied to slaves in a jesting fashion.52 Felicio and his father could
be either literary constructs, or else real slaves who are given appropriate pseudonyms.
Felicios aged appearance might also be fictitious, or at least exaggerated for the
occasion. It is worth at this point considering how old he would be, if he existed, in
A.D. 634, the dramatic date of the letter. Now Seneca jokes that it is Felicios loss of
teeth which makes it likely that he is the pupulus (little boy)53 that his master once
knew. The image Seneca has in mind is the toothless smile of a child whose milk teeth
have fallen out (cf. dentes illi cum maxime cadunt),54 which usually happens at around
five to seven years of age.55 If the estate was acquired no earlier than c. A.D. 13 (cf.
n. 14 above), then Felicio would be no more than in his mid fifties in A.D. 634, but
possibly younger.56 To modern eyes, a person of this age would hardly be expected to
present the appearance that Seneca attributes to Felicio, and we might therefore
conclude that the description is a fiction, although when the living conditions and life
expectancy of the average Roman slave are taken into account, it might not seem
improbable that a slave over fifty could have deteriorated physically to the point of
decrepitude.57 On the other hand, the description is hyperbolically comic, and is
51 Cf. Apuleius Philesitherus (he who loves [amatory] hunting), Met. 9.16, and Philebus,
Met. 8.25 (boy-lover), for meaningful names prefaced by phil-. One thinks here also of the
names of two of the interlocutors in Varros Res Rusticae, Stolo, (Shoot, Sucker) and Scrofa
(Pig), which, although genuine, are clearly selected for the sake of a pun on the agricultural
context; cf. 1.2.9, 2.4.1.
52 Cf. Mart. 3.34 with Watson and Watson (n. 41), 324. This argument is not of course
conclusive, since both are real names. Philositus is rare: H. Solin, Die griechischen Personennamen
in Rom: Ein Namenbuch (Berlin/NewYork, 1982), 1.1656 lists ten examples, of which seven are
definitely slaves or freedmen (cf. H. Solin, Die stadtrmischen Sklavennamen [Stuttgart 1996],
1.235), which could argue either way, i.e. that Philositus is real, or that his name is invented to suit
the context. Felicio is fairly common (23 examples in Solin [1996], 93; there was a slave cobbler
with this name in Domitians court: see M. Charlesworth, Flaviana, JRS 27 [1937], 5462 at
612): it is used of a real delicium on a tombstone, discussed by Setaioli (n. 1), who argues that the
name might have been given to pet slaves who enjoyed the special favour of their owners.
53 The unusual diminutive is found only elsewhere in Catullus 56.5, where it refers to a puer
delicatus, and some (e.g. Costa [n. 6] ad loc.) see a sexual sense here too, but this is unwarranted.
54 Cf. Summers (n. 6), ad loc.; R.M. Gummere, Seneca Epistles 165 (Loeb,
London/Cambridge, MA, 2002), ad loc.
55 Cf. Plaut. Men. 1116 of a seven-year-old: septuennis: nam tunc dentes mihi cadebant
primulum. The toothlessness of old age can be compared with that of a baby (cf. Juv.
10.199200), but the use of cadunt belies that interpretation here. Also, Felicio asks non cognoscis
me? and it is hardly likely, even in fiction, that a person one had known only as a babe-in-arms
would be recognizable 50 years later.
56 This assumes that he was taken up as a delicium (cf. n. 6 above) in the same year that Seneca
built the estate and that he was then at least five years old, but if he had become a delicium a little
later than A.D. 13, and/or had been taken up as an infant, Seneca remembering him as he was at a
slightly older age, then he could be as young as fifty in A.D. 634.
57 Of interest here is a discussion in the jurist Paulus (Dig. 21.1.11: cf. Gell. NA 4.2.12) of
whether the discovery of tooth loss in a slave could be considered valid grounds for returning the

220

PATRICIA AND LINDSAY WATSON

deeply infused with literary sneers against the aged.58 Old people are, for instance,
frequently likened to corpses, as Felicio is here (quis est iste decrepitus et merito
ad ostium admotus? foras enim spectat alienum mortuum).59 Such insulting
comparisons are particularly common in invective against old women, so-called
Vetula-Skoptik, the most striking case being Mart. 3.93.1827, an extended verbal
assault on a uetula as cadauer, but elderly males are also so described: most notably,
Seneca the Elder records a declamatory color used by Latro, in which the use of an old
male slave as a model for a painting of Prometheus, which involved the torture and
consequent death of the slave, was excused on the basis that he was as good as dead in
any case.60 Equally grounded in the thematic armoury of comic taunts against the
elderly is the subject of toothlessness.61 As before, ageing women are the primary
targets of such jeers, but elderly males are by no means exempt. Thus the lecherous
uetulus of Plautus Casina is insultingly described as illius hirqui improbi, edentuli
(550), a similar characterization is applied in the mad scene of the Menaechmi to the
senex of the play (864), while an epigram of Martial (6.74) satirizes the vanity of an
old man who attempts in vain to disguise his loss of hair and teeth. Lastly, the
combined sneer that, in claiming as if in his dotage to have been Senecas one-time
delicium, and in dropping his teeth, Felicio has become like a child (pupulus) again,
once more echoes a theme typically encountered in mockery of the elderly; for it was a
commonplace of such sneers, canonically expressed in Juvenals madidique infantia
nasi, that the old regress mentally and physically to childhood.62
We have demonstrated by a close analysis how Senecas description of his visit to the
estate in sections 13, in terms of its themes, symbolism and diction, is an
embroidered version of real life, overlaid with a patina of literary referentiality. Let us
now briefly consider how the passage functions as a lead-in to the rest of the Epistle.
Seneca begins Epistle 12 with the observation quocumque me uerti, argumenta
senectutis meae uideo, thus announcing by reference to his own situation the central
theme of the Epistle, old age and how one should react to it. Sections 13 take the
form of an autobiographical flashback showing how the author comes to an
acceptance of his own senectus. In it he moves through successive stages of increasing
self-awareness, marked by a movement from anger at the uilicus, who puts it to him
that the dilapidated condition of the buildings is due to old age rather than neglect, to
a wry aside acknowledging that Seneca, like his trees, is growing old, and finally to
mockery at Felicios expense, culminating in the joke about toothlessness, which is
designed to cloak Senecas discomfort at having implicitly to concede that he himself
is growing old. This growth of self-awareness63 is crystallized at the opening of section

slave to the seller as faulty goods: cui dens abest, non est morbosus: magna enim pars hominum
aliquo dente caret neque ideo morbosi sunt: praesertim cum sine dentibus nascimur nec ideo minus
sani sumus donec dentes habeamus alioquin nullus senex sanus esset.
58 A point already suggested by the description of Felicio as decrepitus. The word is
old-fashioned, as Summers (n. 6) notes ad loc., and belongs more to the world of Roman comedy,
where sneers against the elderly are legion, than to current usage. For the comic/invective
tradition of attacks on old people see Parkin (n. 5), 869; cf. Cokayne (n. 5), 16.
59 Cf. Mart. 3.32.2, 10.90.2, and Hor. Epod. 8.12 with Watson (n. 26) ad loc.
60 Sen. Contr. 10.5.17. Cf. also Parkin (n. 5), 64.
61 For sneers against tooth loss in the old, cf. Parkin (n. 5), 834.
62 Juv. 10.199. For the perception of old age as a return to childhood see Parkin (n. 5), index
s.v. second childhood.
63 Senecas attainment of self-awareness about his old age is vividly illustrated in later Epistles,

SENECA AND FELICIO: IMAGERY AND PURPOSE

221

4 by the admission debeo hoc suburbano meo, quod mihi senectus mea, quocumque
aduerteram, apparuit, which then gives rise in the ensuing sections of the Epistle to
philosophical commonplaces on the folly of fretting over the imminence of death and
the necessity of living each day as if it were ones last.
The prooemium to the Epistle, then, is concerned with tracing, by a combination of
circumstantial details and literary symbolism (the symbiosis between the villa/Felicio
and Senecas aged frame), the process whereby Seneca rids himself of his reluctance to
accept that he really has become old. This prepares the ground for the central
hypothesis of the Epistle, that old age is something to be embraced positively, rather
than feared as signifying the approach of death.64
THE PROOEMIUM IN A BROADER CONTEXT
Not only do sections 13 serve as a prooemium to Epistle 12, but they also exemplify
some broader patterns relating to Senecas protreptic techniques in the Epistles as a
whole: (1) the use of supposedly biographical details as a peg on which to hang a
moral or philosophical discussion and (2) self-criticism.
(1) We have seen how Senecas description of a visit to his villa is employed as a
springboard for his reflections on old age. The technique of introducing philosophical
discussion via a personal experience or some circumstance that impinges on Senecas
consciousness is a common one in the Epistles. For instance, an attack of seasickness,
caused by an ill-advised dash across the Bay of Naples, sparks a disquisition on how
blindly we ignore our shortcomings both physical, and, more important, spiritual (Ep.
53), while in Ep. 80 the freedom from interruption occasioned by a boxing match and
mention of the attendant athletic regime suggest, rather factitiously, a lecture on the
need to train the mind, rather than the body.
As in Epistle 12, it is important not to underplay the literary artifice and fictional
nature of these highly circumstantial prooemia. Thus the account in Epistle 53.34 of
the (seasick) Seneca stumbling over the rocks of the shore after being hastily decanted
from his boat is strongly reminiscent of genre-scenes of shipwreck, a literariness
underscored by a facetious reference to the various naufragia suffered by Ulysses in
the Odyssey. Again, the description in 87.14 of the asceticism of the two-day journey
undertaken by Seneca and Maximus, which serves as a platform for a sermon on
the superfluousness of material possessions, reads like a perambulatory version of the
hackneyed theme of the simple life. And sometimes the element of literary contrivance is more obtrusive still. Are we really to believe that Seneca accidentally
stumbled into a noonday intermission at the games without some prior awareness of
the carnage that was to be expected there (casu in meridianum spectaculum incidi, lusus
expectans et sales contra est, 7.3)? Did he, one of the richest men in Rome, really set
up house above the baths (Ep. 56) in order to test the capacity of his philosophical
tranquillity to screen out obtrusive noises,65 or is the colourful and detailed account
of the various noises resounding in the baths, which is occasioned by Senecas living-

where he turns upon his own person the same jokes as he had previously directed against Felicio,
describing himself in 26.1 as decrepitus (inter decrepitos me numera) and repeating in 83.4 the jest
at 12.3 which conflates the dropping of milk teeth with tooth loss in the elderly.
64 Cf. 12.4 complectamur illam (sc. senectutem) et amemus; plena est uoluptatis, si illa scias uti;
in section 6 Seneca rejects the argument molestum est mortem ante oculos habere with the
remark primum ista tam seni ante oculos debet esse quam iuueni.
65 So he claims at the conclusion of the letter 15 experiri et exercere me uolui.

222

PATRICIA AND LINDSAY WATSON

quarters, simply a simulated piece of autobiography, a literary tour de force designed

to carry the message that illa tranquillitas uera est, in quam bona mens explicatur
(56.6), whereas the generality of mankind, bedevilled by greed and ambition, start at
every disturbance? These and similar passages give a strong impression of a kernel of
fact or even pure invention rounded out by a literary circumstantiality which both
satisfies Senecas artistic ambitions66 and subserves his larger didactic purposes as a
proselytizer for Stoicism.
(2) Let us turn now to the subject of self-criticism as a strategy pursued by Seneca
to render his teachings more palatable by displaying his own moral shortcomings.67
Again, the prooemium to Epistle 12 exemplifies a technique which is on display in the
Epistles as a whole. Two notable instances may be pointed to in Epistle 12. The first
involves the bad temper which Seneca exhibits towards the uilicus at the dilapidated
condition of his estate (13). Here Seneca depicts himself acting in a way that
contradicts his own advice given elsewhere in the Epistles and in his earlier work De
ira, that one should not succumb to anger. In Epistle 47, for instance, a disquisition on
the treatment of slaves, he proclaims that a master should in general show himself
hilaris towards his serui (17), while at the conclusion of Epistle 18, apropos of remarks
on ungoverned anger, he suggests that slaves (along with enemies) commonly provoke
this but that the provocation should at all costs be resisted. Moreover, his displaying
anger not only contravenes Stoic principles68 but is especially inappropriate for one
who professes Stoicism, because he depicts himself as acting thus in full awareness of
what he is doing (iratus illi proximam occasionem stomachandi arripio).69
The second example in the Epistle of self-criticism comes in the Felicio episode.
Scholars have often commented on Senecas callousness or indifference in levelling
seemingly malicious jokes at the pathetic old slave.70 In our judgement, however, it is
more productive to view these as a consciously contrived strategy, again involving
self-criticism, which is integral to the didactic effect of the whole.71 The series of ageist
jokes sets Seneca up for a fall when it is revealed by Felicio that the philosopher is
even older than his quondam delicium. The effect is further heightened when, even
after Felicio has revealed his identity, Seneca attempts to mask his embarrassment at
the situation by making a further tasteless sneer about toothlessness. The overall
outcome of the Felicio scene is to portray Seneca in a ludicrous light. What is the

66

Griffin (n. 10), 41819; Wilson (n. 11).

Cf. Griffin (n. 10), 277, 417; Richardson-Hay (n. 9), 3740.
68 The Stoic condemnation of the passions, not least anger, is too well known to need
rehearsing here. See for instance, W.V. Harris, Restraining Rage: The Ideology of Anger Control in
Classical Antiquity (Cambridge, MA/London, 2001), 10418. In Senecas Epistles, see esp. 51.8,
71.37 and 85.12.
69 One might compare Seneca, Thyestes 4889 and 5434, where Thyestes yields to the urgings
respectively of his son Tantalus and Atreus to take up once more the reins of power, acting thus
despite a keen awareness that he is doing an unwise thing, but unable in the end to resist the allure
of kingship.
70 E.g. Wiedemann (n. 1), 1289 Senecas story about how he failed to recognise his old
playmate is an illustration of the callous ancient attitude to old age generally; Parker (n. 1), 242,
n. 56 [Felicios remark] makes no impression on Seneca, despite his professed sympathy for slaves
(Ep. 47 etc); A.J.L. Van Hoof, From Autothanasia to Suicide (London/New York, 1990), 34 the
self-centred Stoic has eyes only for his own sorry fate; Bradley (n. 1), 130 what would Senecas
slave Felicio have thought of the Stoic virtue his master preached?; Setaioli (n. 1), 149. Coleman
(n. 2) says Senecas coarse picture of the toothless old dotard seems callous until we realize that
Seneca sees in him a reflection of himself.
71
Cf. Coleman, cited in previous note, and Henderson (n. 7), 26.
67

SENECA AND FELICIO: IMAGERY AND PURPOSE

223

purpose of such self-mockery? It must be to show in the most colourful way possible
that Seneca himself is not immune to the faults, in this case a lack of self-perception,
which he would seek to eradicate in others. The technique in play here, whereby the
philosopher deliberately places himself in a false or risible position in order to
exemplify his own fallibilities, is repeated in other Letters. For instance, in Ep. 53 the
overblown picture of Senecas seasickness, following a sea-journey unwisely undertaken,72 which sees him clambering over rocks, recontextualizing his enthusiasm for
cold water dips and ascribing Ulysses multiple shipwrecks to an attempt to escape
mal de mer, functions as a comic exemplification of the Epistles moral, that we are all
too readily led to ignore our failings, both physical and spiritual.
Humour deployed by Seneca at his own expense is, as we have just seen, one vehicle
whereby he engages in a salutary self-criticism. In a similar effect, self-criticism can be
implied by an inconsistency between what Seneca says and how he presents himself as
acting. Thus, in the Epistle on the Baths, the comment that he is able to control his
thoughts to such an extent that he can ignore the sounds coming from below is in
conflict with his exceedingly detailed description of the various characters whose
activities he can hear: if he is really immune to such noises why can he describe them
in such vivid detail, and why does he end the letter by saying that he is going to change
lodgings in order not to torture himself (torqueri) further?73 A similar phenomenon is
seen in the famous Epistle on the Games, where Seneca illustrates the moral that the
proficiens should avoid crowds because of their corrupting effect by the example of a
visit on his part to the spectacles during the lunch-time execution of criminals. Here,
the self-criticism is at first more overt: he admits that he returns from the games a
worse person, unable to avoid being contaminated by the other spectators (auarior
redeo, ambitiosior, luxuriosior, immo uero crudelior et inhumanior, quia inter homines
fui, 7.3). Later in the Epistle, however, he describes the wholesale slaughter in the
arena, and in particular the reaction of the crowd, shouting and thirsting for blood,
with vivid disgust (7.45); but though this disgust might seem to represent his reaction
to the events at the time he was witnessing them,74 such a Stoic display of imperviousness to the general fascination with bloodshed is at odds with his earlier statement
that he left the games tainted by those around him. The reader must conclude, then,
that Seneca is superimposing on his account of the spectacle his philosophically
coloured reflections after the event, and that in a more subtle form of self-criticism he
is describing behaviour on the part of the crowd into which he himself was drawn by
their corrupting influence,75 which is why he leaves the Games, as he says himself,
crudelior et inhumanior. We might even speculate that his disgust is all the more

72

53.1 Quid non potest mihi persuaderi, cui persuasum est ut nauigarem?
Ep. 56.15. Cf. Costa (n. 6), 174 ad loc. For a similarly revealing comment at the end of a
moral diatribe, cf. the remark of Thyestes which concludes his laudation of the simple life as
opposed to temporal kingship, immane regnum est posse sine regno pati, where pati notoriously
betrays a hankering for the trappings of power which belies his previous brave words (Sen. Thy.
470).
74 C. Richardson-Hay, Mera Homicidia: a philosopher draws blood Seneca and the gladiatorial games, Prudentia 36 (2004), 87146, at 119 envisages Seneca as standing aloof from the
mob, a lone voice of protest which is not heard above the general din; she fails to address the
inconsistency between this and Senecas earlier admission of being contaminated by his surroundings.
75 This is not, of course, to say that the incident is real, but that such behaviour would be
consistent with the dramatic situation created in the Epistle.
73

224

PATRICIA AND LINDSAY WATSON

virulent for being directed not just towards the crowd but towards his own imbecillitas
for having been unable to withstand their degrading influence.
The specific instances just discussed are part of a broader strategy pursued by
Seneca in the Epistles of underlining at every turn his own fallibilities and
shortcomings as one who is striving imperfectly towards philosophical enlightenment.
Not only does he regularly include himself in the generality of humankind when he
talks of their follies and imperfections (e.g. 59.911 to cite one example among many),
but he also repeatedly foregrounds his own deficiencies as a proficiens who falls
lamentably short of the mark. Examples include 27.1, tu me inquis, mones? Iam enim
te ipse monuisti, iam correxisti? Ideo aliorum emendationi uacas? Non sum tam
improbus, ut curationes aeger obeam, sed tamquam in eodem ualetudinario iaceam, de
communi tecum malo conloquor et remedia communico. Sic itaque me audi, tamquam
mecum loquar; 45.4, sed qualescumque sunt (libri mei), tu illos sic lege, tamquam uerum
quaeram adhuc, non sciam, et contumaciter quaeram; 57.3, non de me nunc tecum
loquor, qui multum ab homine tolerabili, nedum a perfecto absum. This idea comes
particularly to the fore in Epistle 12 in the shape of Senecas bad temper towards the
uilicus which we discussed earlier, instantiating as it does his inability to implement
consistently the principles which he inculcates, a shortcoming which he highlights in
Epistle 75 apropos of a discussion of the three classes of proficientes. Here he says
that he and Lucilius will do well if they are admitted to the third, that is, least
advanced, category of proficiens, a degree of progress which he exemplifies by saying
tertium illud genus extra multa et magna uitia est, sed non extra omnia. Effugit
auaritiam, sed iram adhuc sentit. Against this background, Senecas yielding to anger
in Epistle 12 shows how far he has to go before he reaches the final stage of
philosophical perfection, a condition, according to Stoic thinking, attainable only by
a very few quite remarkable individuals.76
The preceding paragraph on the issue of self-criticism prompts the larger question
of whether Seneca shows this as issuing in some form of moral progression in the
Epistles as a whole. This is too large a topic to explore here, but what can be profitably
noted is that in the penultimate letter of the corpus, 123, Seneca pictures himself
arriving at his Alban villa and finding nothing ready for his arrival. But instead of
exploding into anger at the situation, he accepts it in a spirit of philosophical
imperturbability (mecum enim de hoc ipso loquor, quam nihil sit graue quod leuiter
excipias, quam indignandum nihil [dum nihil] ipse indignando adstruas). It is hard to
escape the suspicion that the incident has been contrived as a positional counterweight to Epistle 12,77 where a similarly unsatisfactory state of affairs in Senecas villa
provoked an outburst of annoyance, suggesting that by the end of the work, Seneca
portrays himself as having attained sufficient self-control and inner quietude not to
respond emotionally to quotidian annoyances.
CONCLUSION
We have, it is hoped, demonstrated, by a close reading of Senecas Epistle 12.13, the
full extent of the literary artifice with which Senecas description of a visit to his villa
is overlaid, and how the events recounted serve as a lead-in to the main theme of the
Epistle, a philosophical discussion of old age. In the second part of the paper, we
showed how this use of quasi-autobiographical material to establish the philosophical
76
77

See Richardson-Hay (n. 9), 38, n. 67.

Thanks to Marcus Wilson for pointing this out and for help in general.

SENECA AND FELICIO: IMAGERY AND PURPOSE

225

keynote of an Epistle is replicated elsewhere in the Epistles. We also discussed a

second Senecan technique exemplified by the prooemium to Epistle 12, the sweetening
of his moral instruction by self-criticism and by the use of humour at his own
expense.
University of Sydney

PATRICIA AND LINDSAY WATSON

patricia.watson@usyd.edu.au
lindsay.watson@usyd.edu.au

Classical Quarterly 59.1 226237 (2009) Printed in Great Britain

doi:10.1017/S00098388090000172

226
THRASYMENNUS WANTON
WEDDING
ROBERT
COWAN

THRASYMENNUS WANTON WEDDING:

ETYMOLOGY, GENRE, AND VIRTUS IN SILIUS
ITALICUS, PUNICA
To set the scene for the battle of Lake Trasimene, Silius presents a unique aetiology
for the lakes name, which he gives as Thrasymennus.1 The beautiful son of Tyrrhenus,
the Lydian colonist and eponymous king of Etruria, catches the eye of the local
nymph Agylle, who, inflamed with passion, carries him down to the depths of the
lake. The nymphs comfort him and, in the capping aetiological couplet, the lake is
named Thrasymennus. This story, almost certainly invented by Silius, is one of several
aetiologies in the poem with a markedly Ovidian character, most closely resembling
the (likewise invented) story of the maiden Pyrene, raped and abandoned by Hercules,
who gave her name to the mountain range.2 As the story of a nympholept, it demands
to be read against the abductions of Hylas by the nymphs and the rape of
Hermaphroditus by Salmacis, but it also has points of comparison with Anna
Perennas union with the river Numicius, of which Silius himself offers a version in
Punica 8.3 Silius regularly adopts and plays with the role of the doctus poeta, both in
the way he manipulates his genuine doctrina and as he wittily invents aetiologies (and
other learned details) in the manner of Callimachus, Propertius and especially Ovid.4
This inventiveness includes a passion for etymologizing, and I shall argue that, in the
case of Thrasymennus, a further etymology lies behind both the name of the lake and
of the beautiful youth. Moreover, this etymology is part of the way in which the story
1 Sil. 5.723. The episode is discussed by P. Asso, Passione eziologica nei Punica di Silio
Italico: Trasimeno, Sagunto, Ercole e i Fabii, Vichiana 2 (1999), 7587, at 758; S. Batino,
Stagnis Thrasymenum opacis: archeologia e miti nella storia di un lago, ArchClass 4 (2003),
41122. I have not been able to locate a copy of M.A. Vinchesi, La vicenda di Trasimeno (Silio
Italico 5, 723) e la fortuna del mito di Ila in et imperiale, in M.P. Pieri (ed.), Percorsi della
memoria 2 (Florence, 2004), 10311. The lake is called Trasumennus at Liv. 22.4.1 (and on 24
other occasions),
at Polyb. 3.82.10, and
at Plu. Fab. 3.1.
2 Ovidian nature of story: R.T. Brure, Color Ovidianus in Silius Punica 17, in N.I. Herescu
(ed.), Ovidiana: rcherches sur Ovide (Paris, 1958), 47599, at 4845; M. Wilson, Ovidian Silius,
Arethusa 37 (2004), 22549, at 229. Pyrene: 3.42041, with A.M. Keith, Engendering Rome:
Women in Latin Epic (Cambridge, 2000), 567; P. Asso, Passione eziologica nei Punica di Silio
Italico: la morte di Pirene, AION(filol) 23 (2001), 21532; A. Augoustakis, Lugendam Formae
Sine Virginitate Reliquit: reading Pyrene and the transformation of landscape in Silius Punica 3,
AJPh 124 (2003), 23557.
3 Hylas: Theoc. 13; Ap.Rhod. 1.120739; Prop. 1.20; V.Fl. 3.5494.57. Hermaphroditus: Ov.
Met. 4.285388. Statius has a similar story about the rape of the youth Lapithaon by the nymph
Dercetis (Theb. 7.297300), which J.J.L. Smolenaars, Statius Thebaid VII: A Commentary
(Leiden, 1994), 145, thinks the poet probably invented. The direction of allusion between Statius
and Silius is notoriously difficult to establish. Anna Perenna: Ov. F. 3.545656; Sil. 8.50201.
4 On Silian doctrina, see esp. A.J. Pomeroy, Silius Italicus as doctus poeta, Ramus 18 (1990),
11939; R. Cowan, Absurdly Scythian Spaniards: Silius, Horace and the Concani, Mnemosyne
59 (2006), 2607. Until recently Silian scholarship was dominated by Quellenforschung, but
among the numerous works devoted to it, see esp. J. Nicol, The Historical and Geographical
Sources Used by Silius Italicus (Blackwell, 1936), and, for a defence of Silius inventiveness, H.-G.
Nesselrath, Zu den Quellen des Silius Italicus, Hermes 114 (1986), 20330. The seminal work on
etymologizing as doctrina is F. Cairns, Tibullus: A Hellenistic Poet at Rome (Cambridge, 1979),
esp. 87110.

THRAS YMENNUS W ANTON WEDDING

227

of Thrasymennus contributes to the wider, and interconnected, themes of generic

propriety and military and moral uirtus in the Punica.
NAMED AND SHAMED: A WANTON WEDDING
The story is full of signposts or markers, which indicate that it is not only aetiological but specifically etymological: nomina seruant (7), dederatque uocabula (11),
nomen (22), dicitur (23).5 Primarily, this establishes the connection between the eponymous figures, Tyrrhenus and Thrasymennus, and, respectively, the land and lake
explicitly named after them. This form of explicit etymology is widely paralleled in
Virgil, Ovid and others, and might be exemplified by Caieta in Aeneid 7 or Croton in
Metamorphoses 15.6 Silius himself provides various instances, especially in the heavily
aetiological catalogue of Roman allies in Punica 8, including Marrus, eponymous
founder of Marruvium, and King Asus, who gives his name to the river Asus and to
the Asili.7 However, with Thrasymennus, the final, etymologizing couplet seems to
overdetermine the derivation (Sil. 5.22-3):
hinc dotale lacus nomen, lateque Hymenaeo
conscia lasciuo Thrasymennus dicitur unda.
From him/this came the lakes dowered name, and, conscious far and wide of the wanton
wedding, the waters are called Thrasymennus.

Silius appears to tell us the same thing twice, that the lake was named after the youth.8
Yet on both occasions he stresses that the naming was inextricably connected with the
wedding of nymph and mortal it was a dowry and it was, in some unspecified way,
associated with knowledge of the wedding a connection which leads us to question
whether this is a simple boy-names-lake etymology. In addition, there is an ambiguity
about hinc. Primarily, the youths dowry is the honour that the lakes name should
come from him, but hinc also suggests that it takes its name from this, that is from
his rape by Agylle.9 If the lakes name is to be derived not only from that of the youth,
5 Asso (n. 1), 77, n. 15. On such etymological markers or signposts, see R. Maltby, The
limits of etymologizing, Aevum(ant) 6 (1993), 25775, at 26870; J.J. OHara, True Names:
Vergil and the Alexandrian Tradition of Etymological Wordplay (Ann Arbor, 1996), 759; A.
Michalopoulos, Ancient Etymologies in Ovids Metamorphoses (Leeds, 2001), 45.
6 Caieta: tu quoque litoribus nostris, Aeneia nutrix, | aeternam moriens famam, Caieta, dedisti; |
et nunc seruat honos sedem tuus, ossaque nomen | Hesperia in magna, si qua est ea gloria, signat.
Virg. A. 7.14, with OHara (n. 5), 183; Croton: nomen tumulati traxit in urbem. Ov. Met. 15.57,
with Michalopoulos (n. 5), 61. On explicit etymologies in the Aeneid, see OHara (n. 5), 735.
7 Catalogue: 8.349621. Marrus: Marruuium ueteris celebratum nomine Marri (8.505; E.M.
Ariemma, Alla vigilia di Canne: commentario al libro VIII dei Punica di Silio Italico [Naples,
2000], 127, compares Serv. ad Virg. A. 7.750: alii Marrubios a rege dictos uelint); Asus: ante, ut
fama docet, tellus possessa Pelasgis, | quis Asus regnator erat fluuioque reliquit | nomen et a sese
populos tum dixit Asilos (8.4435); Delzs emendation Asus for the MSS aesis is marginally
preferable to Alfieris Asis, though both are otherwise unknown.
8 Alternatively, as CQs anonymous reader notes, the conjoining que could be taken as introducing not a repetition but an exegetical expansion in parataxis. This would indeed mean that the
explanation is not given twice, but does not alter the fact that it is a double explanation, deriving
(in both the succinct and the expanded versions) the lakes name from the youth and from the fact
of his abduction and wedding.
9 Both definitions of hinc are covered by OLD s.v. 7b (indicating the origin of a word or
name), but an example, from Glares own selection, of the sense from him is et Capys: hinc
nomen Campanae ducitur urbi (Virg. A. 10.145); for from this (event): redduntur merito debita
uina Ioui. | dicta dies hinc est Vinalia (Ov. F. 4.8989). F. Spaltenstein, Commentaire des Punica de

228

ROBERT COWAN

but also from the story itself, this would be an instance of a less common form of
etymology, which OHara calls a double etymology, whereby a place is (explicitly)
named after a person, whose name is itself (implicitly) etymologized. His examples
include the Porta Carmentalis in Aeneid 8, named explicitly after the nymph Carmentis,
whose name is in turn implicitly etymologized from carmen, prophecy, by the use of
the words fatidicae and cecinit.10 With Thrasymennus the additional detail of the
lakes being conscious of, or complicit in, the wanton wedding, apparently extraneous
to the etymological punch-line which the couplet provides, might lead the reader to
wonder whether it is in fact part of the etymology, or rather of a double etymology;
the lake is explicitly named after the youth, but is his name also implicitly etymologized? The description of the wedding as lasciuus, wanton, is almost oxymoronic,
since Hymenaeus tends to be reserved for proper, ritually correct weddings.11 This
oddity of expression further provokes the reader to look for a reason for its formulation, and that reason is an etymological one. lasciuus Hymenaeus glosses the Greek
, the true words behind Thrasymennus.12
It might be objected that lasciuus is not the closest equivalent of
. Agylles
abduction of Thrasymennus is undeniably bold, and this is made almost explicit when
she casts off one of the commonest antonyms of audacia, her casta pudor.13 But is it
her audacia, her
, which makes the wedding lasciuus?
is, of course, most
easily rendered into Latin by audax, as in Plinys capping of Thucydides, or
and
temerarius, as in Apuleius etymologizing of Thrasyllus name.14 However,
its cognates are also used to describe those who are forward and lacking in sexual
restraint, such as Menanders Thais, Lycophrons Aegiale, and even, in one of Pluis regularly described
tarchs alternative readings, Nausicaa.15 The personified
Silius Italicus: Livres 1 8 (Geneva, 1986), 335, is surely right that, pace Volpilhac (in J. Volpilhac,
P. Miniconi and G. Devallet [edd.], Silius Italicus, La Guerre Punique tome 2, livres VVIII [Paris,
1981], 2), lacus is genitive, rather than nominative: the lake [takes] its dowered name from him
makes little sense, since the dowry ought to go to the groom rather than the lake.
10 Virg. A. 8.33741, with OHara (n. 5), 209. At 224, he likewise shows how 10.198200
explicitly derives Mantua from Manto, and implicitly derives her name from
with
the gloss fatidicae. Double etymologies of this kind, where both are authorized by the text,
should not be confused with the use of the same term for alternative, mutually exclusive etymologies, although I shall argue below that Silius also hints at an alternative derivation for
Thrasymennus.
11 E.g. Catull. 66.11; Ov. Ep. 2.33; V.Fl. 8.149. R. Pichon, De sermone amatorio apud latinos
elegiarum scriptores (Paris, 1902), 165: Hymenaeus semper ad ueras iustasque nuptias refertur.
Cf. the similarly jarring oxymoron of Helens inconcessos hymenaeos at Virg. A. 1.651.
12 Ovid may hint at the first half of this etymology at F. 6.7658, when he warns Augustus not
to repeat Flaminius rashness, calling it temeraria tempora, though his orthography Trasimenus
does not facilitate the derivation. Neither Bmer nor Littlewood comment on this.
13 castumque exuta pudorem (5.15). audacia and pudor opposed: Ter. Phorm. 233; Cic. Verr.
2.5.34, 39, Quinct. 79; Sall. Cat. 3.3.21; Tib. 1.4.1314; Ov. Met. 9.527. See also R. Kaster,
Emotion, Restraint, and Community in Ancient Rome (Oxford, 2005), 556. Cf. D.L. Cairns, Aids
(Oxford, 1993), 151: Tharsos can be a positive antonym to misplaced aids, but it can also be
taken too far, and thus lead to anaidei, the negative antonym of aids.
and
are
paired at Cratin. fr. 377 KA, Dem. Chers. 68.3, Meid. 201.5, Andr. 47.3, Ep. 4.4.8; they very
often appear in scholia as more or less synonyms glossing the same word, e.g. T ad Il. 21.394c.
14 sicut 2 6 0! 8 6 9 (, ita recta ingenia debilitat verecundia,
perversa confirmat audacia (Plin. Ep. 4.7.3, quoting Thuc. 2.40.3); Thrasyllus, praeceps alioquin et
de ipso nomine temerarius (Apul. Met. 8.8). In the other direction, Planudes, in his translation of
the Met., uses
to render ferox (1.758, 9.31), ferus (9.85), fortis (4.652, 6.221), proteruus
(12.234), temerarius (10.545), and uiolentus (8.106).
15 Men. fr. 163 KA, Lyc. 612, Plut. Quomodo adul. 27B.

THRAS YMENNUS W ANTON WEDDING

229

as
, and it is worth noting that the Roman Amor, though he can make his
victims audax, is never so described himself; on the other hand, he is not infrequently
lasciuus.16 lasciuus itself ranges from the neutral playful (OLD s.v. 1) to free from
restraint in sexual matters (OLD s.v. 3), but always retains a sense of going beyond
the acceptable limits, be it of seriousness or sexual propriety.17 That lasciuus has a
semantic range which encompasses boldness is most clearly shown by the author of
the HA life of Hadrian who, in a list of the princeps contradictory qualities, sets the
lasciuus facet of his character in antithesis with his propensity to be a cunctator.18
Pichon even goes so far as to gloss lasciuia, as used by the elegists, as signifying simul
uoluptatem et inpudentem liberamque audaciam, and lasciui as those qui in amore
omnia audent.19 There is, therefore, quite sufficient equivalence between the meanings of lasciuus and
for the etymology to be recognized in such a heavily
signposted couplet, even if the former is not the most obvious translation. Indeed the
, instead of the more obvious audax, is a calculated
choice of lasciuus to calque
, of these hymenaeals was of the
one, designed to underline that the boldness,
wanton, erotic kind and emphatically not of the courageous, martial kind.
NOT BRAVING BUT CLOWNING: THE ETYMOLOGICAL PATH
NOT TAKEN
In privileging lasciuus over audax to render the Greek
, Silius does not only
point the reader towards his preferred, erotic etymology; he indicates the rejected,
martial etymology, whose trace is still present, almost under erasure, as in the
phenomenon which Christopher Ricks has called an anti-pun.20 Critics have noted
that the story is not only Ovidian, but elegiac, bucolic, anything but epic.21 This
incongruity with the surrounding narrative, and especially the great battle which will
dominate the rest of the book, is played out within the story itself.22 Thrasymennus
16
as
: Aristophon fr. 11. KA; Ap. Rhod. 3.687; AP 5.274.1, 7.421.34 (where
he is called simply
); Nonn. 2.223, 33.103, 42.206, 47.267. Love makes people bold: Ov.
Met. 4.96, cf. Lib. Ep. 147.3.5. Amor as lasciuus: Tib. 1.10.57; Ov. Am. 3.1.43; Ars 2.497; cf. the
less clearly personified amores at Hor. Carm. 2.11.7; Ov. Ars 3.27; Mart. Apoph. 187.1. audax
only qualifies amor at Stat. Theb. 4.260, where Parthenopaeus is struck by a paradoxical passion
for metonymous war: audaci Martis percussus amore.
17 TLL VII.2.983.48ff.: modum excedens, imprimis gravitatem, moderationem, continentiam
sim. vel pudicitiam. The sole instance of forms of audax and lasciuus together is unfortunately
unhelpful, Tac. Germ. 24.1 (quamuis audacis lasciuiae pretium est uoluptas spectantium), since the
boldness is in the danger which the sword-dancing entails, and lasciuia is used in its most neutral
sense of sport: A.A. Lund, P. Cornelius Tacitus, Germania (Heidelberg, 1988), 177, compares
Liv. 1.5.2.
18 idem seuerus laetus, comis grauis, lasciuus cunctator, tenax liberalis, simulator
<dissimulator>, saeuus clemens et semper in omnibus uarius. SHA Hadr. 14.11.
19 Pichon (n. 11), 184.
20 Whereas in a pun there are two senses which either get along or quarrel, in an anti-pun
there is only one sense admitted but there is another sense denied admission. So the response is
not this means x (with the possibility even of its meaning y being no part of your response), but
this-means-x-and-doesnt-mean-y, all hyphenated. C. Ricks, The Force of Poetry (Oxford,
1984), 2656. So, in this aition, Thrasymennus-means-wanton-wedding-and-doesnt-mean-boldspirited. I am grateful to Seamus Perry and Carl Schmidt for introducing me to the anti-pun.
21 Elegiac: Asso (n. 1), 77; bucolic: Volpilhac (n. 9), 133.
22 At a late stage of revision, I discovered that A. Augoustakis, Two Greek Names in Silius
Italicus Punica, RhM 148 (2005), 2224, at 2234, also notes the derivation of Thrasymennus
name from
. However, he connects the
, not with the boy himself, but with the
arrogant and immodest Tyrrhenus.

230

ROBERT COWAN

father, Tyrrhenus, is depicted as a culture-hero, who not only gives his name to the
land in which he settles his Lydians, but introduces the trumpet to the locals and thus
ushers in Iron Age civilization23 (5.913):
Lydius huic genitor, Tmoli decus; aequore longo
Maeoniam quondam in Latias aduexerat oras
Tyrrhenus pubem dederatque uocabula terris.
isque insueta tubae monstrauit murmura primus
gentibus et bellis ignaua silentia rupit.
His father was Lydian, the glory of Tmolus; Tyrrhenus had once brought the Maeonian people
by a long sea-journey to Latian shores and given his name to the land. And he was the first to
reveal the unfamiliar sound of the trumpet to the people and to burst the idle silence for war.

Tyrrhenus is thus a sort of militaristic Saturn, or perhaps more aptly an Iron Age
Jupiter from Georgics 1, toughening mankind up, not by the rigours of agriculture,
but by war.24 For the silence which Tyrrhenus trumpet bursts is not peaceful but
ignauus sluggish, lazy, even cowardly.25 His positive and beneficial invention or
introduction of war as a means of improving Italys moral fibre is entirely in keeping
with the Jupiter of the Punica, who, in his own combination of the Georgics 1
theodicy and Aeneid 1 prophecy, explains that he has brought about the Second Punic
War in order to reinvigorate the decadent Romans and spur them to greater glory.26 In
this he parallels the Jupiter of Valerius Flaccus Argonautica, who likewise ushers in
an Iron Age of war (there coupled with sea-faring) in order to raise mankind to the
stars, and, like his Silian counterpart, combines this theodicy with a prophecy of
empire.27 Yet the emphasis on the musical instrument and on the sound of war adds a

23 On Tyrrhenus invention of the trumpet, see Hyg. Fab. 274; D. Briquel, Lorigine lydienne
des trusques (Rome, 1991), 31944.
24 Saturn as culture-hero: Virg. A. 8.31427; Macrob. Sat. 1.7.213; A.O. Lovejoy and
G. Boas, Primitivism and Related Ideas in Antiquity (Baltimore, 1935), 57. Jupiters Iron Age
theodicy: Virg. G. 1.11859. According to Plin. HN 7.201, Tyrrhenus also invented hastae
uelitares and, though the text is uncertain at this point, perhaps the pilum. See Briquel (n. 23),
34568. In the alternative, Capuan Weltanschauung of Teuthras song, the golden age is casta
Saturni saecula patris (Sil. 11.458).
25 Cf. Don Fowlers provocative reading of the arming of the Italians in Aeneid 7: Italy is
recalled from its Saturnian slumber to a Jovian sense of struggle and cultural progress Juno as
culture-hero in her opening of the gates is not simply a matter of parody: there is a sense in which
this is the way progress really is made in human societies. Opening the Gates of War, in H.-P.
Stahl (ed.), Vergils Aeneid: Augustan Epic and Political Context (Swansea, 1998), 15574, at 168
= Roman Constructions (Oxford, 2000), 17392, at 188.
26 3.571629; on greatness through adversity, 57190, esp. 5845: iamque tibi ueniet tempus quo
maxima rerum | nobilior sit Roma malis. On the motif: H. Hommel, Per aspera ad astra, WJA 4
(194950), 15765. On this speech: M. von Albrecht, Silius Italicus. Freiheit und Gebundenheit
rmischer Epik (Amsterdam), 1718; W. Kiel, Das Geschichtsbild des Silius Italicus (Frankfurt
am Main, 1979), 426; W. Schubert, Jupiter in den Epen der Flavierzeit (Frankfurt am Main,
1984), 458, 5580; F. Ahl, M.A. Davis and A. Pomeroy, Silius Italicus, ANRW II.32.4 (1986),
2492561, at 2504; D.C. Feeney, The Gods in Epic (Oxford, 1991), 3056; R. Marks, From
Republic to Empire (Frankfurt am Main, 2005), 1415. This idea, and the corollary that Romes
post-Hannibalic success and prosperity inevitably lead to decline, is central to the Punica.
27 V. Fl. 1.53167, esp. 5656: durum uobis iter et graue caeli | institui. See Schubert (n. 26),
234, 3142; Feeney (n. 26), 3189, 3304; M. Wacht, Juppiters Weltenplan im Epos des Valerius
Flaccus (Stuttgart, 1991); A. Zissos, Sailing and sea-storm in Valerius Flaccus (Argonautica
1.574642): the rhetoric of inundation, in R. Nauta, H.-J. van Dam and J.J.L. Smolenaars (edd.),
Flavian Poetry (Leiden, 2006), 7995.

THRAS YMENNUS W ANTON WEDDING

231

metapoetic dimension to Tyrrhenus revolution: he is not merely introducing war to a

peaceful land; he is an epic poet giving sound to war in place of bucolic silence.
There is a metapoetic aspect to the next line also, as Silius describes the proud
fathers ambitions for his son (5.14): nec modicus uoti natum ad maiora fouebat, And,
immoderate in his prayers, he nurtured his son for greater things. maiora is rather
compressed and even enigmatic, and it is only the connective nec which guides the
reader to understand that these greater things are successes in the wars of the
previous sentence. Well, perhaps not only the nec, for once the connection with
military matters is made, she might think of other contexts where the comparative of
magnus has such a connotation: paulo maiora canemus, nescioquid maius Iliade
nascitur, maior rerum mihi nascitur ordo, | maius opus moueo, certe maiora canebas;28
Tyrrhenus, the ambitious patriarch of a distinguished military family, might well want
his son to follow him into the army, but Tyrrhenus, the martial epic poet, wants his
son to be a martial epic hero. He wants his name to be etymologically connected with
, bold-spirited,29 epithet of the Homeric Heracles.30 Perhaps he even
wants him to be a Hercules-figure, as would be appropriate in an epic aition in the
tradition of Evanders narrative of Cacus, and indeed Hercules features in several
other aitia in the Punica.31 He is doomed to be disappointed.
Agylle is bucolic, elegiac, Ovidian, entirely unepic. Indeed, her elegiac quality is
almost overdetermined, as her passion is described in a bewildering mix of erotic
metaphors (5.15, 1819):
uerum ardens puero castumque exuta pudorem
flore capi iuuenum primaeuo lubrica mentem
nympha nec Idalia lenta incaluisse sagitta.
28 Virg. E. 4.1; Prop. 2.34.66; Virg. A. 7.445; Ov. F. 4.3; cf. Ov. Tr. 2.63. The first and last
strictly refer to panegyric and aetiology, rather than martial epic, but there is considerable
slippage between these higher genres set in antithesis with the lower genres of pastoral and erotic
elegy. On forms of maior as generic markers for epic, and especially martial epic, see F. Bessone
(ed.), Heroidum Epistula XII (Florence, 1997), 356. For examples in Silius, see D.P. Fowler,
Even better than the real thing: a tale of two cities, in J. Elsner (ed.), Art and Text in Roman
Culture (Cambridge, 1996), 5774, at 723 = Roman Constructions (Oxford, 2000), 86107, at
1046, on 6.711, and below on 4.476.
29 So Apollon. Lex. s.v. and Hsch.
697 (
) and, less precisely, A ad.
Il. 5.639 (
); bT ad. Il. 5.639 (
) seems improbable in context,
though more plausible etymologically; perhaps best of all would be the D scholias mildly
paradoxical
. That there was
confusion about the words meaning and etymology contributes to the potential for competing
interpretations here.
30 It occurs only twice, on both occasions in the formula
, in
Tlepolemus vaunt to Sarpedon (Il. 5.639) and Odysseus description of Alcmene in the parade of
women (Od. 11.267); Heubeck (in A. Heubeck and A. Hoekstra, A Commentary on Homers
Odyssey vol. II, books IXXVI [Oxford, 1989], 93) thinks the formula might derive from an earlier
poem about Heracles. One might note that both contexts concern parent-child relationships; the
only other extant occurrence, except for quotations and glosses of Homer, is at Bacchyl. 5.6970,
where Meleager is the bold-hearted grandson of Porthmaon, and is being met in the underworld
by none other than Heracles, all three elements linking back to the Homeric contexts.
31 Parallels with Hercules and Cacus: Asso (n. 1), 75; Hercules also appears in the aitia of
Saguntum (1.27387), Pyrene (3.42041) and the Fabii (6.62736), as well as being a central
figure in the Punica. On Hercules in the Punica, see esp. E.L. Bassett, Hercules and the hero of
the Punica, in L. Wallach (ed.), Classical Tradition: Literary and Historical Studies in Honor of
Harry Caplan (Cornell, 1966), 25873; F. Ripoll, La morale hroque dans les popes latines
dpoque flavienne (Louvain, 1998), 11232; Augoustakis (n. 2). He is also, according to Hyginus,
Tyrrhenus father.

232

ROBERT COWAN

She burns, casts off her modesty like a cloak (and we are perhaps meant to imagine
that that is not all she casts off ), is captured by the flower of the youths beauty,
because she is slippery in her mind, and then she is swift to grow hot (again) as the
result of Cupids arrow.32 A more elegiac figure it is hard to imagine. As such, she not
only steals Tyrrhenus son away, but freezes him forever at the point where he is a puer,
a desired, feminized, objectified figure, who will never grow into an epic uir.33 Like her
Ovidian predecessor, Salmacis, she simultaneously feminizes and elegizes her victim.34
Thrasymennus will never achieve the maiora which his father hopes for; the echo of
Eclogue 4 highlights his failure to match the achievements of that poems precocious
puer, instead sliding back into being a figure like Hylas in Eclogue 6, a neoteric,
bucolic negation of epic (as in Apollonius).35 He will never have the nomen, the
glorious reputation, which all Silian heroes crave.36 Even the name he has will be given
a new meaning, not bold of spirit, but wanton in his marriage. Although the
martial etymology is rejected by omission, rather than being registered and then
32 To give one example each from many, ardeo: Ov. Am. 1.9.33; castus pudor: Ov. Ars 1.100;
casting off restraining emotions, cf. Ov. Ep. 18.57: deposito pariter cum ueste timore; capio: Prop.
1.1.1; flower of youth: Tib. 1.8.47; incaluisse: Ov. Ep. 18.42; sagitta: Tib. 2.1.81. See Pichon (n. 11)
for further examples. lubricus is not a markedly elegiac word, but Varro connects it, and specifically the collocation lubrica mens, with lubet and lubido (L. 6.6.47). R. Ash, Tacitus, Histories
Book II (Cambridge, 2007), 381, commenting on Tacitus use of lubrica to describe Vitellius
fickle fleet at H. 2.101.2, suggests that he might have been struck by Silius description of Agylle.
33 On the puer as the antithesis of the epic uir, as much as the femina, see Ll. Morgan, Childs
play. Ovid and his critics, JRS 93 (2003), 6691. In Valerius Hylas episode, Juno, though the
protectress of Jason, because of her overriding hostility to Hercules, similarly threatens to derail
the epic mission of the Argonauts and thus the generic status of the poem, as Jupiter ironically
reminds her (4.78): sic Iuno ducem fouet anxia curis | Aesonium, sic arma uiro sociosque ministrat?
(on which see Feeney [n. 26], 324). The deified Hylas urges Hercules to cease his elegiac lament
(quid, pater, in uanos absumis tempora questus? 4.25) and to return to his epic ways (in duris haut
umquam defice, 35). On the generic connotations of allusions to arma uirumque, see also E.L.
Bassett, Silius Punica 6.153, CPh 54 (1959), 1034, at 1314; A. Bloch, Arma virumque als
heroisches Leitmotiv, MH 27 (1970), 20611; D. Hershkowitz, Patterns of madness in Statius
Thebaid, JRS 85 (1995), 5264, at 63; B.W. Boyd, Arms and the man: wordplay and the
catasterism of Chiron in Ovid Fasti 5, AJPh 122 (2001), 6780; for questus and queror as markers
for elegy (erotic and funereal), see J. Ingleheart, Ovid Tristia 1.2: high drama on the high seas,
G&R 53 (2006), 7391, at 84, with further references, ancient and modern; for epic duritia and
elegiac mollitia, see S.E. Hinds, The Metamorphoses of Persephone (Cambridge, 1987), 214; D.F.
Kennedy, The Arts of Love (Cambridge, 1993), 314.
34 On the Salmacis episode as a conflict of epic and elegy, see I. Jouteur, Jeux de genre dans les
Mtamorphoses dOvide (Louvain, 2001), 27280. On the feminization of Hermaphroditus, see
M. Robinson, Salmacis and Hermaphroditus: when two become one, CQ 42 (1999), 21223;
P.B. Salzman-Mitchell, A Web of Fantasies: Gaze, Image, and Gender in Ovids Metamorphoses
(Columbus, 2005), 325, 1603. On Valerius Flaccus allusion to this episode to suggest Medeas
threat to Jasons manliness and epic status, see T. Stover, Confronting Medea. Genre, gender, and
allusion in the Argonautica of Valerius Flaccus, CPh 98 (2003), 12347, at 12733.
35 puer: Virg. E. 4.8, 18, 60, 62; Hylas: E. 6.433. I am grateful to Denis Feeney for suggesting
the further implications of the passages intertextuality with the Eclogues. For an argument that
Valerius Flaccus reclaims Hylas for epic in polemical response to Propertius elegizing of him, see
M.A.J. Heerink, Going a step further: Valerius Flaccus metapoetical reading of Propertius
Hylas, CQ 57 (2007), 60620.
36 nomen as glory: OLD s.v. 11, Sil. 2.699, 3.595, 4.37, 6.462, 7.600, 10.71, 13.98. There is
frequent slippage between this sense and OLD 10 The name (of a person) as famed and OLD 2
The name (of a people considered to have individual existence), as at 10.5012: sed iuueni, ne
sim tibi longior, hinc est / et genus et clara memorandum uirgine nomen. Cloelias fame, her gentile
name and the fame of that name are all won by her courageous deeds. The contrast with
Thrasymennus is the more pointed, since his nomen (name) is immortalized, but, as the passive
figure of an erotic escapade rather than an active military hero, he has no nomen (glory).

THRAS YMENNUS W ANTON WEDDING

233

dismissed, the adjudication between competing etymologies is typical of such aitia,

and especially of Ovids technique in the Fasti.37 Tyrrhenus tried to nurture (fouebat)
his son for greater things, but now the Naiads stroke (fouere) him to comfort him.38
Even the glossing of
as lasciuus associates it, not with epic audacia, but with
the playful, childish Ovidian poetics which Llewelyn Morgan has shown to be so
disruptive of the grown-up decorum of epic.39 Elegy, forever railing against the
of war, has its revenge on epic Tyrrhenus.40
IN THE NAME OF THE SON: THRASYMENNUS AND THE PUNICA
Silius aetiologies have often been seen merely as decoration, demonstrations of his
doctrina to satisfy the tastes of Flavian readers. Asso concludes that the story of
Thrasymennus d un colore e una vivacit sorprendenti e spezza la monotonia della
narrazione, while Wilson vividly argues that the Punicas Ovidian elements as a whole
invite readers to look at the view through the window and not at their digital
watches.41 However, they are able to produce more subtle, serious and wide-ranging
effects. Keith and Augoustakis, respectively, have shown how the aetiological story of
Pyrene exposes the violence that underwrites the assimilation of the female to the
topography of epic, and dramatizes Hannibals failure to follow the positive aspects
of his Herculean model successfully.42 Batino suggestively explores the mythical
implications of the Thrasymennus narrative, with its parallels in Etruscan myth, and
its allegorical interpretation as a transition either into another phase of life or from
life to death.43 We have seen that there is a generic resonance to the story, fore37 Ov. F. 1.31936, 2.44950, 47580, 3.83946, 4.612, 8590, 5.1110, 6.1100, a subset of
the list of all alternative aetiologies in J.F. Miller, The Fasti and Hellenistic didactic: Ovids
variant aetiologies, Arethusa 25 (1992), 1131, at 12, n. 6. In 5.1110, as in our case, the choice
between etymologies has a generic resonance: It is not only a question of an aetiological choice,
for each Muse brings a different type of poetic discourse to bear on the argument, and each has a
bent towards the tradition of a different literary genre. (A. Barchiesi, Discordant Muses,
PCPhS 37 [1991], 121, at 14).
38 solatae uiridi penitus fouere sub antro | Naides amplexus undosaque regna trementem, 5.201.
The detail recalls Theoc. 13.534, except that there the same nymphs both snatch and comfort
Hylas. The erotic overtones of fouere (cf. Tib. 1.6.6, with J. Adams, The Latin Sexual Vocabulary
[Baltimore, 1983], 208) are further in contrast with Tyrrhenus paternal nurturing. If the scene
also reminds the reader of Aristaeus being comforted and entertained by the water-nymphs at
Virg. G. 4.35985, she might note Thrasymennus failure to become a comparable culture hero.
39 Morgan (n. 33), passim and esp. 6975.
40 Schetliasmos of
of war: Tib. 1.3.478, 1.10.1; Prop. 4.3.1920 (occidat,
immerita qui carpsit ab arbore uallum | et struxit querulas rauca per ossa tubas) is particularly pertinent to Tyrrhenus. On curses of inventors more generally, see F. Leo, Plautinische Forschungen
(Berlin, 1895), 1514, and in love elegy, R. Mller, Motivkatalog der rmischen Elegie (Zrich,
1952), 21.
41 Asso (n. 1), 78; Wilson (n. 2), 238. I agree with Wilson that Virgilian source material is
always co-opted into a process of generic transgression (247), but would argue that there is an
ideological dimension to such generic transgression.
42 Keith (n. 2), 57; Augoustakis (n. 2), 254. Despite the constant effort of the Carthaginian
general to imitate Hercules, Hannibal is portrayed as a follower of an erroneous model, whose
darkest traits Silius has carefully underscored.
43 [S]i pu facilmente intuire come il mito del rapimento di Trasimeno, con la transizione
spaziale dal dominio terrestre a quello acquatico, possa aver assunto gi in fasi molto antiche il
significato di una perfetta metafora per indicare il passaggio ad unaltra dimensione, che sia
quella del regno di Ade o il passaggio ad una nuova fase della vita. (Batino [n. 1], 416). Hylas
rape in Theocr. 13 has likewise been interpreted as a sort of death by C. Segal, Poetry and Myth in
Ancient Pastoral (Princeton, 1981), 5461.

234

ROBERT COWAN

grounded by the conflict of etymologies, which dramatizes the tension within the
poem between its epic nature and elements which try to draw it towards lower genres.
I would like to conclude by arguing that, as I have already intimated, this generic
conflict is, on a symbolic level, a microcosm of the poems wider concerns about
moral revival and decline. Tyrrhenus success in epicizing and militarizing the Etruscans, and his failure to continue that level of generic and moral integrity into the next
generation, are played out on a larger scale throughout the Punica.
The positive transition from slothful peace to active, glorious, Iron Age war, which
we have seen Jupiter give as his reason for starting the conflict, is initiated when Fama,
emphatically placed as the incipit of Punica 4, spreads the news that Hannibal has
crossed the Alps. It is hard not to think of the Virgilian Fama, whose appearance in
another book 4 led (albeit at three removes) to Aeneas abandonment of the slothful
life of the elegiac lover in favour of resuming his epic mission. Generic propriety and
divinely ordained duty go hand in hand.44 So it is in Punica 4, where the polishing of
rusty spears in the manner of Virgils Latins not only marks their return to the
militarism which made Rome great, but restores the epic propriety of the poem, as
Mars ciet arma uirosque, summons arms and men at the same time as he summons
epic.45 As with Aeneas, the generic conflict is not mere literary cleverness but a
means of expressing the conflict between the value systems which epic and its antitheses represent. Fama, Mars and even the unwitting Hannibal here serve the will of
Jupiter to make Rome great through war and through epic, and in this they are a
precise parallel for Tyrrhenus.
Yet the Punica is not only a poem of cultural progress and renovation; it is also a
poem of moral and political degeneration and decline. Elegiac elements are always
threatening to defeat the epic, and simultaneously the forces of decadence threaten to
demoralize those whose martial valour would otherwise win them glory. Disturbingly,
a distinctly elegiac Venus and her Cupidines save Rome by corrupting Hannibal and
his Carthaginians during their sojourn in Capua, simultaneously a contrast to the
hardiness of third-century B.C. Rome and a prolepsis of the decadence of Domitianic
Rome.46 Scipio, replaying the Prodican Hercules, rejects the personified Voluptas in
favour of Virtus, but the former, bearing an uncanny resemblance to the Silian Venus,
flounces off with the chilling prophecy that her day will come, which, in Silius moral
pessimism, it of course already has.47 The echoes of Ovids similar encounter with
44 Love between Dido and Aeneas runs counter to the will of fate, but also contradicts the
generic canons of epic since it represents, on more levels than one, an intrusion of materials
outside and not provided for in the epic code (e.g., erotic-elegiac, erotic-tragic). The dialectical
overcoming of the deviant Carthaginian episode ends up being therefore victory for epic no less
than for Fate. (A. Barchiesi, Speaking Volumes. Narrative and Intertext in Ovid and Other Latin
poets [London, 2001], 131).
45 4.11. For arma uirumque as metonymy for epic, see under n. 33 above. Silius is particularly
perhaps excessively fond of plays on the formula: 1.132, 241, 364, 519, 2.675, 3.526, 4.98, 253,
5.325, 6.6, 7.8, 8.272, 661, 9.100, 597, 10.505, 554, 12.1689, 189, 17.102, 279, 4423, 516.
46 11.385431. On this scene, see E. Burck, Silius Italicus. Hannibal in Capua und die
Rckeroberung der Stadt durch die Rmer (Mainz, 1984), 224; G.O. Hutchinson, Latin Literature
from Seneca to Juvenal (Oxford, 1993), 2037; A. Barchiesi, Genealogie letterarie nell epica
imperiale, in Lhistoire littraire immanente dans la posie latine (Vanduvres-Genve, 2001),
31554, at 33642; R. Cowan, Indivisible Cities: Mirrors of Rome in Silius Italicus (Oxford, forthcoming). The enervating winter sojourn spent in love and luxury which threatens the generic
purity of the epic has two key antecedents in Apollonius Argonauts visit to Lemnos, on which
see Feeney (n. 26), 3224, and of course Aeneas in Carthage (see n. 44 above).
47
Scipio in biuiis: 15.18129; Voluptas parting shot: 1237. On the scene, see esp. Ahl-Davis-

THRAS YMENNUS W ANTON WEDDING

235

Tragoedia and Elegeia in Amores 3.1 re-inforce the sense that both moral and generic
levels are in play, that Scipio has chosen both uirtus and epic, but that vice and elegy
will triumph in the end. Agylle, the elegiac puella who turns a prospective epic hero
into an eternal puer delicatus, the man of the bold spirit into the boy with the wanton
wedding, embodies the generic and moral threat to the epic, military ideal which the
Punica depicts as the only way to the stars, as opposed to the bottom of a lake.
This threat is, of course, to the future, to succeeding generations. The Punica is
pre-occupied with descent familial, ethnic and literary and how the past shapes the
present and the present the future.48 The conflict between Rome and Carthage is
based on their descent from Trojans and Didos Tyrians, and what happens in the
course of the Second Punic War will have (with the benefit of hindsight) an immense
impact on the future of Rome down to Silius own Domitianic era. We have already
seen how Tyrrhenus hopes for his son, that he will be a great warrior and (which
comes to the same thing) an epic hero, are frustrated. This symbolizes, on a small
scale, the poems grander concerns about the future of Rome, whose martial, epic
qualities can too easily be corrupted into elegiac decadence. However, it gains even
greater resonance from being read in context, following immediately from the end of
Punica 4, when Hannibal refuses the Carthaginian demand (fomented by his enemy
Hanno) to have his son sacrificed in a tophet.49 Hannibal, like Tyrrhenus, has great
ambitions for his son, ambitions that he will be a great warrior and, of course, a great
epic hero.50 This wish is most clearly encapsulated in one lapidary line (4.814): at puer
armorum et belli seruabitur heres. Silius favoured collocation of arma uirumque is here
adapted to puer armorum: Hannibals puer will, as his father hopes, grow up to be an
epic uir, who can be coupled with arma to produce the epic formula. He will be the
heres, the heir, both literal and literary, to the arma and the bellum, the incipit of the
Aeneid and the title of the Bellum Ciuile. Hardie has shown how Hannibal fails in his
attempt to substitute a sacrifice of the Roman dead at Trasimene for that of his son,
so that his hopes that his son will take his place as a great leader of his people will
come to nothing.51 The last time we saw him with his wife and child, he was cast as a
Hector, keeping them away from battle.52 Like Hector, his hopes for his Astyanax, that
he will be a great warrior like his father, are vain. We might add that his hopes, like
Tyrrhenus, and indeed Hectors, that his son would take his place as the hero of an
epic poem also come to nothing.
This contrasts sharply with the young Scipio, a puer who will even in the course
of the poem develop into an epic uir, rather than being frozen in unfulfilled
childhood. Indeed, even in Book 4, he moves from being an Ascanius, marked out for
Pomeroy (n. 26), 25534; M. Fucecchi, Lo spettacolo delle virt nel giovane eroe predestinato:
analisi della figura di Scipione in Silio Italico, Maia 45 (1993), 1748, at 424; Marks (n. 26),
14861.
48 P.R. Hardie, The Epic Successors of Virgil (Cambridge, 1993), esp. 88119; Ripoll (n. 31),
4863; Cowan (n. 4).
49 4.763829, with Hannibals speech at 80929. On this scene, see esp. R.T. Brure, Silius
Italicus Punica 3.62162 and 4.763822, CPh 47 (1952), 21927, at 21922; M. Fucecchi, Irarum
proles: un figlio di Annibale nei Punica di Silio Italico, Maia 44 (1992), 4554.
50 Again, I find myself partially anticipated by Augoustakis (n. 22), who notes the parallel
between Tyrrhenus and Hannibal, but not the generic dimension.
51 Hardie (n. 48), 501.
52 3.62162, with Brure (n. 49), 2223; von Albrecht (n. 26), 1467; D.W.T. Vessey, The dupe
of destiny: Hannibal in Silius, Punica III, CPh 77 (1982), 3205, at 3245; Ahl-Davis-Pomeroy
(n. 26), 25134; Ripoll (n. 31), 669. This scene also evokes Pompeys leave-taking of Cornelia at
Luc. 5.722815, but the latter resonance does not extend to the debate over child-sacrifice.

236

ROBERT COWAN

greatness by the omen of the dove which lands on his helmet before the battle of
Ticinus, to a blend of the maturing Ascanius who kills Numanus Remulus and the
Aeneas who rescues his father from Troy, as Scipio does his at the battle of Trebia.53
The augur Liger, interpreting the omen correctly, addresses the puer as one who will
have a maius Carthagine nomen (4.130), on one level, of course, Africanus, but also a
fame, for all his current childish years, appropriate to the higher genre of epic; it will
certainly not be the elegiac name wanton wedding.54 He is marked out again, in
Jupiters instructions to his son Mars,55 as a puer, but one who already entrusts to his
tender right hand battles, proelia, half of that other recurrent metonymy of epic, the
reges et proelia which Apollo prevented VirgilTityrus from singing; with Mars help,
even, but that he audeat, not
he might dare his first-fruits of battle, to be bold,
lasciuiat.56 When Scipio, Aeneas-like, has performed the pius feat of rescuing his
father, Mars praises him (macte) in the manner of the Apollo of Aeneid 9, marking
Ascanius for a great future. He addresses him as care puer, but the echo of Evanders
farewell to Pallas is a contrastive one: Scipio will not be a doomed youth, deflowered
by death, but will live on to a fulfilled epic future. For Mars predicts that he will go on
to greater things, militarily and generically, in the way that Thrasymennus will not:
adhuc maiora supersunt.57 In this respect Scipio resembles the boy Hannibal, whom
Hamilcar (unlike Tyrrhenus with his son) successfully inspires with epic passion
(1.80): Romanum seuit puerili in pectore bellum. Almost poignantly the epic which the
great Barca hopes he has sown in his boys breast is called Bellum Romanum, a title
which can only be given by a Carthaginian victor, to distinguish one successful war
from the others; just as the Greeks fought the Persian Wars, so Rome victorious will
brand this conflict with the name of the defeated, the Second Punic War, the
Hannibalic War, and its epic will be, not Romana, but Punica.
For Scipio, not Hannibal, will be the ultimate victor of this epic. His superiority to
Hannibal, to Thrasymennus, and to the other youths who are among the poems
defeated, is manifested in two interlinked ways. Firstly, his destined victory means
that, though he starts as a puer, he swiftly matures to become an epic uir, and is never
frozen in the passive state of an ermenos, as the feminized, eroticized object of a
desiring gaze. Building on the work of Fowler and others, Reed has shown how, in the
eroticized deaths of figures like Virgils Euryalus, Pallas and even Turnus, the shadow
of a feminine persona figuratively registers the loss of their adult male potential; the
erotic light that falls on them in a sense confirms their now permanent status as boys
and we are shown the failure of these fallen warriors to continue national or family
lines.58 Thrasymennus rape by Agylle, a sort of thanatized love rather than eroticized
53 Omen: 4.10119, with Ahl-Davis-Pomeroy (n. 26), 2544; Fucecchi (n. 47), 203; Marks
(n. 26), 1639.
54 Denis Feeney suggests per litteras that there may be a further play on his later title of Maior,
to distinguish him from Scipio (Aemilianus) Africanus Minor.
55 The generational continuity of martial and epic excellence is further marked by Jupiter
addressing Mars (notably not Mercury or Apollo) as nate (4.420).
56 4.4256. Cf. Virg. E. 6.35; for reges et proelia as metonymy for martial epic, see e.g.
G. Williams, Banished Voices. Readings in Ovids Exile Poetry (Cambridge, 1994), 2932. On this
scene: Ahl-Davis-Pomeroy (n. 26), 2544; Fucecchi (n. 47), 235; Marks (n. 26), 38, 21819.
57 Mars speech: 4.4727; Apollo: Virg. A. 9.6414; Evander: ibid. 8.581. On this scene:
Ahl-Davis-Pomeroy (n. 26), 2545; Fucecchi (n. 47), 279; Hardie (n. 48), 97; Ripoll (n. 31), 1823;
Marks (n. 26), 37; and esp. (noting the generic relevance) S. Casali, The poet at war: Ennius on
the field in Siliuss Punica, Arethusa 39 (2006), 56993, at 589,
58 J.D. Reed, Virgils Gaze (Princeton, 2007), 23, 40. Cf. D.P. Fowler, Vergil on killing virgins,
in M. Whitby, P. Hardie and M. Whitby (edd.), Homo Viator: Classical Essays for John Bramble

THRAS YMENNUS W ANTON WEDDING

237

death, as Batino has shown, and his reduction from prospective epic uir to the object
of her desiring gaze render him one of the poems defeated: emasculated, productive
of no generational continuity, leaving a name to be sure, but only one commemorating his shameful marriage.59 Hannibal too will finally become the object of the
Roman peoples desiring gaze, even if it is only as an absent presence, when the imago
of him in defeat, fleeing on the plain, holds everyones eyes and minds.60 By then,
Scipio will be addressed, not as care puer, but as inuicte parens (17.651).
Secondly, Scipios choice of Virtus over Voluptas, both personified and abstracted,
of epic over elegy, sets him up as the morally and generically superior figure, who,
within the epic economy of the Punica, can fulfil the moral and generic plan of
Jupiter. Thrasymennus succumbs to Agylle, Hannibal to Venus at Capua, but Scipio
resists temptations both allegorical and corporeal. For, when, after the capture of
Nova Carthago, he sends back untouched the betrothed virgin offered to him as a
spoil of war, Laelius praises his continence, his refusal to veer into the territory of
both decadent voluptuary and elegiac lover, in terms which explicitly compare him (to
his advantage) with epic heroes qua epic heroes (15.275-6): cedat tibi gloria lausque |
magnorum heroum celebrataque carmine uirtus.61 Indeed, even the heroes of the Iliad,
Agamemnon, Achilles, and all the other inhabitants of tents on the Trojan plain,
succumbed to this weakness; for weakness it was they violated their treaty of
alliance femineo amore, primarily out of love for women such as Chryseis and Briseis,
but also out of a feminine love, an incontinent, enervating, elegiac passion, which
leads to moral ruin and disaster, as it did when Camilla conceived her feminine desire
for spoils.62 It is Scipios ability to resist such temptation, unlike Thrasymennus and
Hannibal, which marks him as the epic hero envisaged by Jupiter and by Tyrrhenus,
the puer who will grow into a uir and gain a greater name, not wanton wedding, but
Africanus.63
Balliol College, Oxford

ROBERT COWAN
bob.cowan@balliol.ox.ac.uk

(Bristol, 1987), 18598; for the motif in Silius contemporary, Statius: C. Jamset, Death-loration:
the eroticization of death in the Thebaid, G&R 51 (2004), 95104.
59 See n. 43 above.
60 sed non ulla magis mentesque oculosque tenebat, | quam uisa Hannibalis campis fugientis
imago (17.6434). Hardie (n. 48), 389, emphasizes the triumphs substitution of imagines for the
real Hannibal and Scipio.
61 On this scene, see Ripoll (n. 31), 3523, 4634; Marks (n. 26), 2389. Cf. Ahl-DavisPomeroy (n. 26), 2554, who see it, on the contrary, as an exception to their depiction of an already
corrupted Scipio.
62 15.27732. Camilla: Virg. A. 11.782.
63 I am very grateful to Denis Feeney and Elly Cowan for their helpful comments on drafts of
this article.

Classical Quarterly 59.1 238246 (2009) Printed in Great Britain

doi:10.1017/S00098388090000184

238
IN THECHARLES
WAKE OF McNELIS
LATONA

IN THE WAKE OF LATONA:

THETIS AT STATIUS, ACHILLEID 1.198216
At Thetis undisonis per noctem in rupibus astans,
quae nato secreta velit, quibus abdere terris
destinet, huc illuc divisa mente volutat.
200
proxima, sed studiis multum Mavortia, Thrace;
nec Macetum gens dura placet laudumque daturi
Cecropidae stimulos; nimium opportuna carinis
Sestos Abydenique sinus. placet ire per artas
Cycladas; hic spretae Myconosque humilisque Seriphos
205
et Lemnos non aequa viris atque hospita Delos
gentibus
qualis vicino volucris iam sedula partu
iamque timens, qua fronde domum suspendat inanem;
providet hic ventos, hic anxia cogitat angues,
hic homines: tandem dubiae placet umbra, novisque
215
vix stetit in ramis et protinus arbor amatur. (Achilleid 1.198207; 21216)

Standing on the shores of Thessaly, Thetis deliberates about where to hide Achilles so
as to prevent him from joining the expedition to Troy. Statius geography has troubled
critics in various ways. D.R. Shackleton Bailey, for example, asks whether Statius
mind was in Lemnos, with memories of Hypsipyle when the poet claimed that
Thrace was nearest to Mt Pelion (Ach. 1.201); on Statius inclusion of Lemnos among
the Cyclades (Ach. 1.2056), Shackleton Bailey accurately states that Lemnos is not,
in fact, located there, but he does not elaborate.1 Similarly, O.A.W. Dilke, seemingly in
an effort to explain geographic mistakes, suggests that Statius had never visited
Greece and thus his use of Greek topography is second-hand and imprecise.2
Examples of such criticisms could easily be multiplied.3 Absent a recourse to textual
emendation, for which the mansucripts offer no grounds, we are left with limited
explanations for the apparent oddities. One is to continue to suggest (or imply) that
Statius was misinformed, incompetent or both; another, that this geography serves
literary ends. This paper argues for the latter.
The ancients themselves knew that poets represented lands that were, in fact,
distant from one another as contiguous and ones that were contiguous as distant.4
Moreover, the constitution of the Cycladic islands in particular was hardly fixed in
antiquity: Strabo, for example, claims that twelve islands were initially classified
1

D.R. Shackleton Bailey, Statius: Thebaid, Books 812; Achilleid (Cambridge, 2003), 328.
O.A.W. Dilke, Statius: Achilleid (Cambridge, 1954), 129.
3 E.g. J. Mheust, Stace, Achillide (Paris, 1971), 15. See P.J. Heslin, The Transvestite Achilles:
Gender and Genre in Statius Achilleid (Cambridge, 2005), 135, n. 64 for further bibliography.
4 Strabo 1.2.20 comments on geographic oddities in, for instance, Sophocles and the prologue
of Euripides Bacchae. Geographies in Roman poetry were also challenging: the Propempticon
Pollionis, the work of the meticulous poet Cinna, contains an itinerary for the area around
Actium that provoked questions from ancients; cf. A.S. Hollis, Fragments of Roman Poetry c.
60BCAD20 (Oxford, 2006), 256. W. Kroll, Studien zum Verstndnis der rmischen Literatur
(Stuttgart, 1924), 293307 is a standard discussion of geographical knowledge (or the lack
thereof ) in Roman poetry; cf. too R. Mayer, Geography and Roman Poets, G&R 33 (1986),
4754.
2

IN THE W AKE OF LATONA

239

among the Cyclades but that additional ones were added later (10.5.3); yet Pliny
names thirteen islands that make up the chain (HN 4.658). In part this loose classification is likely to stem from the fact that the Homeric poems an authoritative
though hardly indisputable source of geographic knowledge do not mention the
Cyclades. Not surprisingly, poets such as Pindar (fr. 33d), Callimachus (Del. 36) and
Virgil (Aen. 3.756) variously utilized the shifting position of Delos and the instability
of the Cyclades for their own poetic ends.5 Indeed, in the Achilleid itself, Peter Heslin
has already demonstrated that Statius creatively uses Delos as a foil to Scyros in order
to enhance an understanding of the doomed attempts of Thetis to save Achilles.6
Such a marked literary history suggests that Statius topography warrants closer
analysis.
The geography of this particular passage of the Achilleid reworks passages from
Homeric and Callimachean hymns, Euripides, Virgil and Ovid in ways that, for the
most part, create a Thetis who is modelled upon Latona, the mother of Apollo and
Diana who was forced to travel around the Aegean in order to find a place to give
birth. Though Thetis is not about to deliver a child, the parallels between the two
divine mothers highlight constraints against which the Nereid contends. Yet, as is
suggested by the simile of the mother bird that immediately follows Thetis deliberations (Ach. 1.21216), the mother of a child who is fated to a mors immatura must
and does think and act differently from Latona, the mother of the eternal youth
Apollo. Indeed, the analogy between Thetis and Latona ultimately draws a strong
contrast between the mortal Achilles and the divine Apollo. Statius Aegean
topography and the subsequent simile thus show that Thetis persistently looks to yet
alters Letos path where necessary in order to try to keep Achilles safe.
In trying to circumvent the heroic narrative(s) by which Achilles dies and gains his
epic fame, Thetis is certainly doomed to failure. But whereas she often seems inept
and a poor reader of the literary tradition,7 in this case her assessment of the best
location to keep Achilles safe shows her to be a cautious and alert reader of past
poetry. It is just unfortunate for Thetis that her very awareness of the literary past also
emphasizes her experiential understanding of fundamental distinctions between
humans and the gods in the world of epic.
I. PROXIMA THRACE
Statius statement that Mt Pelion is next to Thrace seems patently wrong in terms of
actual distance, but this juxtaposition of Pelion and Thrace is paralleled in the
Homeric Hymn to Apollo. Specifically, a catalogue of the wanderings undertaken by
the pregnant Leto places Mt Pelion next to Thrace, or at least to Thracian Mt Athos
|
and Samos (h.Ap. 334
). These seemingly distant lands are literally juxtaposed, suggesting that for a
goddess, spatial arrangements may be viewed from an entirely different perspective.
5 For Pindar and Delos, see I. Rutherford, Pindars Paeans: A Reading of the Fragments with a
Survey of the Genre (Oxford, 2001), 24352, 36472; for Callimachus, see M. Giuseppetti, LInno
a Delo di Callimaco: aspetti del mito e dello stile (Diss., Universit degli Studi Roma Tre, 2008);
M. Depew, Delian Hymns and Callimachean Allusion, HSCP 98 (1998), 15582; J.
Nishimura-Jensen, Unstable geographies: the moving landscape in Apollonius Argonautica and
Callimachus Hymn to Delos, TAPA 130 (2000), 28999; A. Barchiesi, Immovable Delos: Aeneid
3.7398 and the Hymns of Callimachus, CQ 44 (1994), 43843 discusses the Virgilian island.
6 Heslin (n. 3), 1347.
7 Heslin (n. 3), 1059.

240

CHARLES McNELIS

Moreover, these locations appear in the itinerary that Leto follows as she moves from
Crete up the western side of the Aegean to Thrace to the eastern Aegean and then
down towards the Cyclades. The hymns description of her circular pattern of movement establishes a kind of proximity between the western Aegean (and specifically
Pelion) and Thrace.8 Thetis mental movements, then, recall Letos actual wanderings.
Statius reworking of the Homeric hymn not only links Thetis with Latona but it
also situates his poetic geography in a tradition of topographical incongruity. After
all, the itinerary of Letos travels in the hymn is hardly precise, since Scyros, for
example, is mentioned among islands that are found in the eastern Aegean (h.Ap. 35).9
Additionally, the famous division of the hymn into Delian and Pythian halves
dispenses with cartographic tidyness.10 Clearly, then, the hymn itself is predicated
upon geographic displacement, and Statius claim about the proximity of Thrace to
Pelion thus locates his scene in a tradition that operates with a fluid sense of
geography.
For this specific expression of Thraces location, however, Statius most of all
develops Virgilian geographical interests. At Aen 3.13, Virgil describes Thrace as terra
procul vastis colitur Mavortia campis. The adverb procul often calls attention to long
distances (OLD s.v. 2), and in the context of Thrace it may seem natural to take it in
that sense. In fact, later in the poem Virgil represents Thrace as the prototypically

distant land (Aen. 12.3356 ultima Thraca; cf. Hom. Il. 10.434
). But since Thrace is Aeneas first stop after he left Troy,11 it is actually not all
that far from the fallen kingdom and thus procul may signify here a separation that is
not of great distance. Indeed, on this very passage, Servius explains that procul means
not very far (non valde longe), and elsewhere he notes that the word may mean both
near and far.12 Virgils procul thus points to the difficulty of defining Thraces
position in absolute terms: is it near or far?
The Virgilian question is revisited by Statius in Achilleid 1.201. The adjective
proxima establishes that Thrace is quite literally closest, yet the position of Thrace in
the final foot of the hexameter is as distant as possible from its modifier that begins
the verse. The hyperbaton and the adjective simultaneously suggest both proximity
and separation. Moreover, Statius description of Thrace as Mavortia (Ach. 1.201), an
adjective that is first attested in Virgil, echoes the account of Thrace at Aen. 3.13.13
Statius proxima, then, seems to pick up on and respond to Virgils geographic play
with procul. That is, he offers that, to judge from the travels of Odysseus and Aeneas,
8 Nor is the geography of the hymn exceptional since even in the Iliad, Thraces separation
from mainland Greece is variable. At Il. 14.227, for example, the land that is defined as Thrace
seems further west than does the very eastern territory that is dubbed Thracian at Il. 2.84450;
see R. Janko, The Iliad: A Commentary. Volume IV. (Cambridge, 1994), 1867.
9 Heslin (n. 3), 1356 has shown the importance of the hymns treatment of Scyros for Statius
own account.
10 See J.J. OHara, Inconsistency in Roman Epic (Cambridge, 2007), 1516 and 26 on the
notorious geographical problems in the Homeric Hymn to Apollo that are then reworked by
Callimachus Hymn to Artemis. He suggests that the geographical incongruities may be seen as
part of a tradition and literary technique rather than a point of confusion.
11 Thrace is also the first stop for Odysseus (Od. 9.39).
12 Cf. Servius comments on Aen. 5.124 (where he also points to the example of Ecl. 6.16) and
Aen. 6.10. N. Horsfall, Virgil: Aeneid 3 (Leiden, 2006), 53 notes the need to read the sense of
distance from Aeneas Trojan perspective.
13 Statius engagement with the geography of Aeneid 3 is discussed more generally by
H. Kuerschner, P. Papinius Statius quibus in Achilleide componenda usus esse videatur fontibus
(Marburg, 1907), 624.

IN THE W AKE OF LATONA

241

Thrace is a short distance from Troy.14 It is no wonder, then, that Thetis avoids Thrace
at the very moment when she is trying to decide how to keep her son from going to
fight and die at Troy.
Statius startling reversal of Thraces conventional epic epithet suggests that this
topographic account is relative. That is, it depends upon Thetis perception of
distance as it relates to Achilles and Troy. Nonetheless, her viewpoint is not without
justification since it evokes Letos wandering in the Homeric hymn. Both points
indicate that Thetis conception of Aegean geography depends upon the literary past.
II. STATIUS CYCLADES
After she rules out Macedon, Athens, Abydus and Sestus (Ach. 1.2024), Thetis
focusses upon the Cyclades as a hiding place for her son. Once again, earlier poetic
geography illuminates her decisions and motivations. Initially, the configuration of
the islands seems odd: Myconos, Seriphos and Delos are all recognizable as parts of
the Cycladic chain, but Lemnos, located in the north Aegean, is not even close to
these islands. However, as mentioned earlier, the identity of the islands that actually
made up the chain was debated in antiquity (Strabo 10.5.3), so it seems best to
consider this strange arrangement of the Cyclades in light of literary history.
Two points emerge from doing so. First, Statius geography calls to mind the
passage from Euripides Trojan Women in which Poseidon assures Athena that he will
rouse the waters around Delos, Myconos, Scyros and Lemnos in order to destroy the
Greeks upon their return home from the Trojan War (8990). Euripides does not
classify these islands as the Cyclades, but none the less the islands he mentions are
relevant for Statius passage. After all, Statius mentions three of the Euripidean
islands, and the fourth (Scyros) need not be cited since it is where Thetis ultimately
chooses to hide her son. In addition to the topographical similarities, however, the
context of the Euripidean passage matters because it points to a crucial difference
between the goddesses. Specifically, Athena is able to convince Neptune to rouse a
storm and to overwhelm the Greeks who sailed to Troy, whereas Thetis has failed to
convince Neptune to create a storm that would destroy the Greek fleet and thus save
Achilles (Ach. 1.6195). Moreover, Statius Neptune is like the Euripidean sea-god (to
say nothing of Poseidon in the Odyssey) in that he looks to the Greek return trip not
the outbound one from Troy as the moment to destroy the fleet. The timing of the
divine assault upon the Greek fleet is crucial for Thetis hope to save Achilles, but
the Euripidean geography, evoked by Statius in part through the unusual inclusion of
Lemnos among other Cycladic islands, highlights the futility of her attempts. She is
truly a weakened and desperate goddess.15
The second point is that Statius description of the Cyclades also situates the
passage in an Ovidian tradition that creates and exploits geographic incongruities.
The collocation of Seriphos and Myconos, for example, alludes to Ovids catalogue of
14 Indeed, Thrace is the European land closest to Troy. The problematic proximity of Europe
and Asia is brought into sharper relief when Thetis decides to avoid Sestus and Abydus (Ach.
1.204). These two sites were conventionally seen as the shortest crossing point of the Hellespont
and thus the closest point of contact between Europe and Asia (e.g. Hdt. 7.44; Ov. Ep. 18.127).
When discussing the Trojan War earlier in the poem (Ach. 1.82), Statius, echoing a Virgilian
expression (Aen. 10.901 quae causa fuit consurgere in arma | Europamque Asiamque), tellingly
views it as the collision of Europe and Asia. Thetis decision to pass over certain sites, then, has
real justification. For more on the boundaries between the continents, see D. Feeney, Tenui latens
discrimine: spotting the differences in Statius Achilleid, MD 52 (2004), 85105, at 1015.
15 On Thetis lack of power, see Heslin (n. 3), 1604.

242

CHARLES McNELIS

islands that join Minos in his attack on Athens (Met. 7.4634),16 and perhaps the
martial heritage of these islands prompts Thetis to avoid them. However that may be,
Statius handling of Myconos is particularly revelatory. Two of the three appearances
of Myconos in extant Latin literature before Statius come from Virgil (Aen. 3.76
Mycono e celsa) and from this section of Ovids catalogue (Met. 7.463 humilem
Myconem). The conflicting descriptions of the islands height is it celsa or humilis?
that the Augustan poets offer are acknowledged by Statius juxtaposition of Myconos
with the epithet humilis,17 but he ultimately sidesteps the issue by transferring the
Ovidian adjective for the island to Seriphos.18 Nonetheless, Statius diction calls attention to the (deliberate) confusion and contradictions that appear in poetic decriptions
of the Cyclades.
Statius designation of Delos, the eventual centre of the island group, as hospita
introduces even greater topographical confusion. The phrase hospita Delos replicates
the language used by the Ovidian Niobe when she, the mother of fourteen children,
rants against her townspeople for worshipping Latona, who has only two offspring.
In the midst of her speech, Niobe imagines that the wandering Delos said to Latona
that they are counterparts in that one wanders on land, the other on sea (Met. 6.1901
hospita tu terris erras, ego dixit in undis | instabilemque locum Delos dedit). The idea
of movement, suggested both by erras and hospita,19 creates the expectation that
Statius Delos will be like Ovids wanderer and that Thetis once again envisions the
Aegean in relation to Latonas movements.
In terms of understanding Statius geography, a wandering Delos matters because
its movement precludes a fixed constitution of the Cyclades.20 Disturbed boundaries
caused by a floating Delos are already manifest in Callimachus hymn, in which the
island moves south from Euboea, the northernmost boundaries of the Cyclades,
before reaching its fixed location in the centre of the chain (Del. 1968). According to
such a principle, Heslin argues, Scyros, which had created geographic problems as far
back as the Homeric hymn, could be positioned among Statius Cyclades.21 Perhaps,
then, Statius representation of the Aegean is so distorted that even more northern
locales such as Lemnos may be represented among the Cyclades.22 If so, the inclusion
of Lemnos among the Cyclades not only echoes the passage from Euripides Trojan
16 Ovids lengthier arrangement focusses on the Cyclades but, like Statius, also includes
outlying islands such as Astypalaia (Met. 7.463; one of the Sporades: cf. Strabo 10.5.14) and
Peparethos (Met. 7.470; not classified among Strabos list of the Cyclades at 9.5.16). G.L.
Huxley, Arne Sithonis, CQ 32 (1982), 15961 fruitfully discusses Ovids inclusive geography.
17 Bmer (on Met. 7.464) discusses the Ovidian correction.
18 Another allusion to Ovid is the connective -que placed at the end of humilis, resulting in a
line ending that matches Ovids mention of Seriphos (Met. 7.464 planamque Seriphon).
19 For hospita, cf. TLL 6.3.3032.2432; 3033.1727.
20 As Heslin (n. 3), 136 remarks on Thetis speech to Scyros, In the world of the mythical past
that Statius is describing, the geography of the Aegean was still somewhat unstable .
21 Heslin (n. 3), 135; it is intriguing that Scyros seems out of place as well at Catullus 64.35,
where the manuscript reading Scyros has been emended to Cieros. The poem, of course, takes
Thetis marriage as its subject.
22 It should also be noted that, as she does with other locations that pose some sort of problem
for her interests, Thetis is wise to avoid Lemnos. The island is hostile to men (Ach. 1.206), a claim
that refers to the myth that with the exception of Hypsipyle the Lemnian women killed their
husbands and indeed all of the men on the island. Moreover, at this point in the narrative, Thetis
plan to dress her son as a girl has only been hinted at obliquely (Ach. 1.1412), so it seems that the
Achilleids crucial question about whether gender and identity are essential or constructed (see,
e.g., Heslin [n. 3], 2945) is addressed right here, but Thetis knows that her son cannot escape his
biology.

IN THE W AKE OF LATONA

243

Woman, but it may also build upon the literary history of Delos shifting position in
the Aegean.
The Ovidian Delos, however, ultimately proves to be only part of Statius picture.
At the start of Achilleid 1.207, Statius gentibus, emphatically placed in enjambment,
clarifies that hospita functions attributively, meaning welcoming. In other words, the
initial reading and understanding of Delos as a wanderer was deceptively
provisional.23 In fact, rather than referring to travel and movement, hospita actually
marks Delos stability; for, as Callimachus indicates, only after it is rooted does Delos
become open to people from all over the world (Del. 27882; cf. Ach. 1.2067 hospita
gentibus). And there is good reason to think of Callimachus account here since
after it is fixed, Callimachus Delos celebrates the fact that it will no longer wander
). The Callimachean
is virtually
(Del. 4.273
glossed by the initial reading of Statius hospita, and even after the enjambment
changes the meaning of hospita, the meaning of the Callimachean phrase is still
evoked since Statius implicitly makes the point that Delos is no longer a wanderer.
Statius diction and artful enjambment thus activate a second, distinct account about
the (in)stability of Delos.
Ultimately, these two competing mythic versions about Delos are closely related.
After all, before Statius, Ovid had exploited the polyvalence of hospita to suggest that
Delos is on the move, the very opposite point that Callimachus had made through
. And while Statius redefinition of hospita seems to stabilize Delos, the
shifting, Ovidian Delos remains in play. In fact, later in the poem Thetis, having
deposited Achilles on Scyros, leaves the island and again uses Niobes language to
prophesy that if Scyros protects her son, it will not be surpassed in fame by floating
Delos (Ach. 1.388 nec instabili fama superabere Delo ~ Met. 6.191 instabilemque locum
Delos dedit). Just as the literary tradition has it both ways, so too does the Achilleid:
Delos is unstable, both fixed and moving.
A wandering Delos, as discussed above, allows Statius to avoid a fixed Aegean
topography and to recall the wanderings of Latona. But a fixed Delos helps to explain
why Thetis would avoid the island. That is, the rooting of the island coincides with a
new name: what had been called Asteria, according to Callimachus, becomes Delos
(Del. 40; 514).24 This alteration of the name of the island seems to be a Callimachean
(Del. 53), which Roman readers treated
innovation that puns upon the word
, clear or conspicuous.25 Because
as an alpha-privativized form of the Greek
23 For this kind of reading of enjambment in Latin poetry, see, e.g., S. Wheeler, A Discourse of
Wonders: Audience and Performance in Ovids Metamorphoses (Philadelphia, 1999), 1113;
R. Jenkyns, Labor Improbus, CQ 43 (1993), 2438, at 247; W. Batstone, On the surface of the
Georgics, Arethusa 21 (1988), 22745, at 230. The phenomenon continually recurs in poetic
language: at 2289 Batstone discusses critical approaches to the phenomenon in Shakespeare;
for Dantes Inferno, cf. J. Ferrante, A poetics of chaos and harmony, in R. Jacoff (ed.), The Cambridge Companion to Dante (Cambridge, 20072), 181200, at 187: The primary effect of
enjambment in the Inferno is to force the reader to rethink a previous notion.
24 Giuseppetti (n. 5), 456 discusses the temporal relationship of the change of Delos name in
Callimachus and how his hymn relates to versions in Pindar and the Homeric hymn.
25 J.J. OHara, True Names: Vergil and the Alexandrian Tradition of Etymological Wordplay
(Ann Arbor, 1995), 1656 cites Roman interpretations of the etymology. Giuseppetti (n. 5), 24
points out Callimachus innovation with the islands name (cf. 46 as well for his discussion of
Asteria, a name that itself probably involves etymological play upon the adjective
[firm,
fixed] and the alpha-privative). K. Ukleja, Der Delos-Hymnus des Kallimachos innerhalb seines
Hymnensextetts (Mnster, 2005), 12947 and W.H. Mineur, Callimachus: Hymn to Delos. Introduction and Commentary (Leiden, 1984), 75 also discuss Callimachus treatment of the islands
name.

244

CHARLES McNELIS

of the aid it provided to Latona, the Callimachean island turns from a floating, hardto-find land mass to one that is fixed and clear to see. For Thetis interest in keeping
her son away from the Greek army and going to Troy, an open and accessible island
should be (and actually is) the last place she should consider.26 Thetis, then, once
again astutely draws upon the literary past particularly as it relates to Latona in
deciding where to hide her son.27
III. MOTHERS AND SONS
The simile of the mother bird that follows Thetis deliberations reinforces this
etymological play concerning Delos and Latona (Ach. 1.21216). Though the content
of the simile recalls Achilles famous claim at Il. 9.3235 that he is like a mother bird
toiling on behalf of its young, Statius keeps Thetis at the fore. Indeed the points of
contact between Thetis and the mother bird are numerous (timens, 1.213 ~ timidae,
1.211; anxia, 1.194, 214; the anticipation of threats to offspring (providet, 1.214 ~
video, 1.34); both settle on places that are pleasing (placet, 1.211, 215).28 In this regard,
it is intriguing that the mother bird ultimately decides that the umbra ought to be her
nesting place (1.215 tandem dubiae placet umbra). The birds choice may afford
protection, but the fact that the semantic range of umbra includes darkness and
obscurity pointedly counters the etymology of conspicuous Delos. The obscuring
shade pleases the mother bird,29 and in this regard she shares even more common
ground with Thetis, who sensibly did not want to hide her son on an island that is
defined by visibility and clarity.30
A deep poetic heritage underlies these maternal motivations. In particular, the
language and imagery of light and darkness engage with central points of the Iliad. In
Iliad 18, for instance, Achilles claims that by refusing to fight he was not a light for
Patroclus or the other Greeks (Il. 18.102). Moreover, when he subsequently enters
battle after having received his armour from Thetis, four similes describe Achilles
luminous brilliance (Il. 19.37599). The implications are clear: the martial Achilles is
26 It is from this perspective of visibility that Thetis must avoid Delos, which otherwise would
seem fitting since death and warfare are banished from the island (Call. Del. 2767). But the
point is that Thetis must shield Achilles to prevent his discovery, which will lead to his death at
Troy.
27 For potential connections between Callimachus Asteria and earlier mythic accounts of
Thetis (a point that would make Statius engagement with Delos even richer), see Giuseppetti
(n. 5), 436.
28 G. Aric, L Achilleide di Stazio, ANRW 2.32.5 (1986), 292564, at 2937 notes the similes
relevance for Thetis. As noted by D. Mendelsohn, Empty nest, abandoned cave: maternal
anxiety in Achilleid 1, Classical Antiquity 9 (1990), 295308, at 301, Thetis is also connected with
the lioness whose cub Achilles had taken (Ach 1.16870). A triangulated relationship between
Thetis, Latona and the bird may also be suggested by the fact that the mother bird, like Latona, is
seeking a place to give birth.
29 There is no strong reason to correlate the name Scyros with umbra, but one may wonder
whether Statius puns not upon
but
in relation to the island: T Il.23.3323, for
example, quotes Aristarchus and connects the Greek verb
to shade with
.
Moreover, Hesychius claims that Philetas connected
with groves and woods, a place in
which one would naturally find shade (cf. K.Spanoudakis, Philetas of Cos [Leiden, 2002], 3302).
See also P. Chantraine, Dictionnaire tymologique de la langue grecque: histoire des mots (Paris,
1983), 1019.
30 D. Kozk, The dawn of Achilles. Light imagery in the Iliad and Statius Achilleid, Acta
Ant. Hung. 47 (2007), 36985, at 3767 convincingly discusses from another perspective the veil
of darkness that surrounds Achilles time on Scyros and its lifting when Achilles joins the Greek
army.

IN THE W AKE OF LATONA

245

characterized by light, whereas his withdrawal and separation from the Greek army
are a period of darkness. By avoiding Delos Thetis seeks safe obscurity for her son as
opposed to providing the weapons with which he becomes resplendent. She thus
reverses her Iliadic actions and attempts to thwart essential imagery of the Homeric
poem.
Statius contrast between light and darkness as it pertains to Delos also entails
distinctions between the human Achilles and the divine Apollo. The comparison
between the two is set up at the beginning of the scene in Thessaly, when a simile
modelled upon a Virgilian comparison in which Apollo returns to his mothers Delos
(cf. Aen. 4.1439 qualis Lyciam Apollo; cf. Aen. 4.144 Delum maternam)
describes Achilles return to Thetis in terms of Apollo returning to his mother (Ach.
1.1656 qualis Lycia venator Apollo | cum redit et saevis permutat plectra pharetris).31
This explicit comparison is reinforced by the recurring similarities between Thetis and
Latona in subsequent verses.
The relationship between Achilles and Apollo, however, is qualified in large part
through the imagery of light and darkness. Whereas the Homeric Achilles could be
compared to the sun (or Hyperion; cf. Il. 19.398), Statius Thetis declares in
frustration that if Achilles father had been divine, then her son would have been a
great star (Ach. 1.2535 aetheriis ego te complexa tenerem | sidus grande plagis,
magnique puerpera caeli | nil humiles Parcas terrenaque fata vererer). Thetis use of
puerpera, an exceedingly rare word in epic, certainly recalls Ovids Latona (Met.
6.337) and furthers the comparison between the mothers and sons. In addition, while
the contra-factual life that Thetis fantasizes about for Achilles sounds Jovian,32 sidus
may also describe the sun (e.g. Tib. 2.1.47; Ov. Met. 1.424). By Statius day the
connection between Apollo and the sun had been made through consistent puns upon
his appellation Phoebus (e.g. Lucr. 6.1197; Aen. 8.720),33 so Thetis fantasy raises the
possibility that the comparison between the sons operates on an astral level. But of
course Thetis is aware that her son will not achieve such a position in the universe, and
in this sense Apollos association with the sun distinguishes the god from Achilles.
It is in the earthly manifestations of their brilliance that the crucial distinction
between Achilles and Apollo is made. Earlier literature had naturally transferred to
Delos the brilliance and visibility that is associated with Apollo (e.g. Arist. fr. 488
Rose; Pliny, HN 4.66). Yet it is striking that one of the earliest descriptions of Delos
radiance is predicated upon Achilles own brilliance. Specifically, Pindar describes the
), and his use of the rare adjective surely
island as far-shining (fr.33c.6
recalls the cognate form that describes Achilles highly visible tomb in the Odyssey
). In its characterization of sites sacred to Achilles and Apollo,
(Od. 24.83
literary history had linked the god and hero in terms of their visibility and brightness.
But for Statius Thetis, Achilles is not Apollo because, instead of being an ethereal,
31 Heslin (n. 3), 1834 and 253 well discusses the simile in light of connections between Apollo
and Achilles, both of whom are paradigms of ephebic beauty; cf. Ach. 1.1678; Il. 2.6734.
Achilles is a virtual doublet of Apollo elsewhere: see, e.g., W. Burkert, Apellai und Apollon,
RhM 118 (1975), 121, at 19; R. Rabel, Apollo as a model for Achilles in the Iliad, AJP 111
(1990), 42940.
32 Magni caeli, for example, evokes Jupiter, whose association with the sky was widespread
(e.g. Var. LL 5.67 quod Iovis Iuno coniunx et is Caelum; Virg. Aen. 3.171; Ov. Met. 13.707). See
too Kozk (n. 31), 382, for a valuable discussion of the cosmic imagery of these verses in relation
to Jupiter.
33 The connection between Apollo and the sun also appears in Greek as well; cf. Euripides
Phaethon (TGF 781.1113).

246

CHARLES McNELIS

radiant divinity, like all mortals he is earth-bound and destined to die.34 Indeed, it is
through his death that Achilles equals Apollos brilliance.
Thetis knows all too well that her son is like Apollo and she like Latona in many
ways except in the crucial one. Consequently, at a pivotal moment of Achilles life she
tries to prevent his death so as to deny Achilles the tomb that gives him a brilliance to
match Apollos. By fighting against the imagery of light and darkness that
underscores the heroic version of Achilles life and death, Thetis plays her maternal
role differently from the way she does in the Iliad. While it has rightly become
accepted that Statius turns to Ovidian epic to confront the Homeric tradition,35 here
Horatian lyric specifically the ode about Sybaris and Lydia that ends with the
equation of the hiding lover to that of Achilles on Scyros (Carm. 1.8) provides an
explicit model.36 In addition, reconceptions of the kind of heroism embraced by the
Iliadic Achilles appear even in Homeric epic, and perhaps no statement is more
powerful than one offered by the shade of Achilles himself when he states that he
would rather be alive as a labourer than be king of the dead (Od. 11.48991).37 In the
Achilleid, Thetis determined efforts to gauge the relative safety of various places in
the Aegean poignantly show that she comprehends before his death what the poetic
tradition has her son learn only after it.
Georgetown University

CHARLES M C NELIS
cam72@georgetown.edu

34 In this regard the phrase that describes Lemnos (Ach. 1.206 non aequa viris) significantly
points towards Achilles human, that is, mortal, nature. In fact, the rare litotes non aequa is
evocative of epitaphs or funerary contexts, and in this particular metrical sedes and in relation to
Achilles, the phrase echoes Odysseus lament over the death of the hero in Ovids epic (Met.
13.1312 quem quoniam non aequa mihi vobisque negarunt | fata).
35 E.g. S. Koster, Liebe und Krieg in der Achilleis des Statius, Wrzburger Jahrbcher fr die
Altertumswissenschaft n.F. 5 (1979), 189208; G. Rosati, Stazio: Achilleide (Milan, 1994), 2533;
A. Barchiesi, La guerra di Troia non avr luogo: il proemio dell Achilleide di Stazio, AION 18
(1996), 4562; S. Hinds, Allusion and Intertext: Dynamics of Appropriation in Roman Poetry
(Cambridge, 1998), 13642.
36 Horace claims that Sybaris shirks military and athletic training in the sunny Campus
Martius (1.8.34 apricum campum solis) because of his love for Lydia. Sybaris avoidance of
the outdoors implies his effeminancy (cf. NH on Carm. 1.8.4), but as M. Lowrie points out
(Horaces Narrative Odes [Oxford, 1997], 120), Horaces emphasis on sunlight also evokes
Pindars sunny Olympia (O. 1.111). The heroic activity that takes place at the radiant Olympic
games contrasts with Pindars assignation of a shadowy existence to those who lack the courage
to undertake heroic feats (O. 1.815; cf.
). The lack of activity on the part of Horaces
Sybaris is thus markedly antithetical to glorious, masculine endeavours, and this polarity between
masculine military/athletic training and feminine hiding is enhanced by the opposition between
light and darkness.
37 For discussion of the Odyssean retrospective comment upon the heroic values of the
Iliad, see A. Edwards, Achilles in the Odyssey: Ideologies of Heroism in the Homeric Epic
(Knigstein/Ts., 1985), 502.

Classical Quarterly 59.1 247262 (2009) Printed in Great Britain

doi:10.1017/S00098388090000196

247
CHAEREAS
REVISITED
KOEN
DE TEMMERMAN

CHAEREAS REVISITED. RHETORICAL CONTROL

IN CHARITONS IDEAL NOVEL CALLIRHOE
INTRODUCTION
In ancient novel scholarship, the distinction between the ideal Greek novel and its
comicrealistic Latin counterpart has been, and still is, highly influential. It originates
with R. Heinzes thesis that Petronius Satyricon develops from a literary genre
parodying idealistic features in the Greek novels.1 Despite the contributions of
scholars warning against applying this dichotomy too rigidly,2 the distinction remains
a commonly accepted tool to classify novelistic literature.3 In this paper I will focus on
the characterization of the male protagonist in Charitons Callirhoe, the oldest of the
so-called ideal novels.4 My reading of this character will suggest that Charitons
position within the ideal genre should be reassessed, and that consequently the overall
distinction between ideal and realistic novels is a generalization that does not take
into account the actual complexity of one of the oldest representatives of the genre.
The distinction between ideal Greek and realistic Latin novelistic literature is
largely informed by the divergent depiction of character in both sub-genres. Whereas
the Latin novel adopts realistic and sexually explicit character portrayal, scholars
have underlined the idealizing aspects in the characterization of protagonists in the
Greek novel. Their beauty invests them with a godlike appearance, and their nobility
) generates loftiness of character that sharply distinguishes them from other,
(
less noble, characters in the story.5 Scholars have emphasized the unreal atmosphere
surrounding this characterization.6 E. Rohdes view that the protagonists in the Greek
novel are seelenlose Gestalten and Gliederpuppen, invested with a leere und
1

R. Heinze, Petron under der griechische Roman, Hermes 34 (1899), 494519.

Cf., e.g., F. Wehrli, Einheit und Vorgeschichte der griechisch-rmischen Romanliteratur,
MH 22 (1965), 13354. Recently, A. Barchiesi, Romanzo greco, romanzo latino: problemi e
prospettive della ricerca attuale, in L. Graverini, W. Keulen and A. Barchiesi, Il romanzo antico.
Forme, testi, problemi (Rome, 2006), 193218 points to a number of less idealistic elements in
various Greek novels.
3 Cf., e.g., N. Holzberg, Der antike Roman. Eine Einfhrung (Darmstadt, 20063), 59138,
classifying the texts under the headings of Der idealisierende Roman: ltere Texte (5979), Der
komisch-realistische Roman (80111) and Der idealisierende Roman: Jngere Texte (11238).
4 Callirhoe was probably written within one or two decades either side of A.D. 50. This view is
defended by B.P. Reardon, Chariton, in G. Schmeling (ed.), The Novel in the Ancient World
(Leiden, 1996), 30935, at 317, and E.L. Bowie, The chronology of the earlier Greek novels since
B.E. Perry: revisions and precisions, Ancient Narrative 2 (2002), 4763, at 57, who dates
Chariton between A.D. 41 and 62. For an overview of different accounts of the dating of
Chariton, see S.D. Smith, Greek Identity and the Athenian Past in Chariton: The Romance of
Empire (Groningen, 2007), 2, n. 4.
5 E.g. F. Napolitano, Leucippe nel Romanzo di Achille Tazio, Annali della Facolt di Lettere
e Filosofia della Universit di Napoli 26 (19834), 85101, at 86: characters are fortemente
stilizzati because of a forte processo di idealizzazione (with reference made to the beauty of the
protagonists).
6 E.g. D. Del Corno, Anzia e le altre, Atti del II Convegno Internazionale. La donna nel mondo
antico (Turin, 1989), 7584, at 84: Certo, la protagonista del romanzo greco una figura ideale,
per non dire irreale: come gi la stessa eccezionalit dei suoi connotati fisici e anagrafici esplicitamente ammette.
2

248

KOEN DE TEMMERMAN

leblose Idealitt, has never been substantially contested.7 Scholars who point to the
presence of some psychological realism in the protagonists characterization do not
go any further than making short, occasional suggestions. A. Lesky, for example,
conjectures that the influence of rhetorical school curricula on the novelists must
have led, at least for the more gifted, to a greater profundity of the intellectual
processes and to a more refined elaboration of psychological details,8 but he does not
develop this suggestion in any detail. Common opinion still has it that realistic
character depiction is to be looked for (to a certain extent, at least) primarily in the
characterization of minor characters.9
Yet, it has often been pointed out that psychologically realistic detail plays a more
important role in Charitons novel than in the other extant novels.10 It is telling,
however, that only the characterization of the minor characters and occasionally
of Callirhoe have been adduced to support this thesis.11 The characterization of
Charitons male protagonist Chaereas, on the other hand, has been largely neglected.
According to J. Helms, the author of the only systematic study on characterization in
Chariton up to now, it is not even worthwhile to look for any realistic detail in
Chaereas characterization in the first place: There is such a dearth of realistic
detail that a discussion of realism in the case of Chaereas would be unprofitable and,
therefore, it has not been considered further.12
Long before Helms, J. Dunlop wrote in his History of Fiction that Chariton was the
first writer of romance who succeeded in depicting an interesting male character.13
Since Dunlop assumed, like Rohde later, that Chariton was the latest of all Greek
novelists, Chariton is, in his view, not only the first, but also the only novelist applying
psychological characterization. Significantly, however, Dunlops statement does not
refer to Charitons protagonist Chaereas, but to the Milesian antagonist Dionysius.
The limited attention that Chaereas characterization has received centres primarily
upon his assimilation with pre-existing character types. It has been pointed out, for
example, that Chaereas is associated, merely by his name, with the character type of
the adulescens, often bearing the same name in New Comedy.14 This character type is
hot-tempered and passionate, and Chaereas name might be read as an implicit
prolepsis of his uncontrolled outburst in the first book of the novel. Furthermore,
Chaereas assimilation with epic and tragic heroes and the inversions of and divergences from these paradigms have also received some attention.15 So far, however,
7

E. Rohde, Der griechische Roman und seine Vorlufer (Leipzig, 1914), 4767.
A. Lesky, A History of Greek Literature (trans. J. Willis and C. de Heer; orig. Geschichte der
griechischen Literatur [Munich, 19632]) (London, 1966), 859.
9 Cf. B.P. Reardon, The Form of Greek Romance (Princeton, 1991), 26; Holzberg (n. 3), 63.
10 G. Schmeling, Chariton (New York, 1974), 1578; A. Billault, Aspects du roman de
Chariton, IL 33 (1981), 20511, at 206.
11 Cf. Rohde (n. 7), 430; B.P. Reardon, Theme, structure and narrative in Chariton, YClS 27
(1982), 127, at 13. J. Helms, Character Portrayal in the Romance of Chariton (The Hague/Paris,
1966), 12746 includes a small chapter on realism in small details. Although Helms credits
Charitons heroine with a couple of individual traits accentuated by the use of realistic and
picturesque details (129), he traces realism especially in the portrayal of the minor dramatis
personae (128). On psychologically realistic features in Callirhoes characterization, cf. K. De
Temmerman, Blushing beauty. Characterizing blushes in Charitons Callirhoe, Mnemosyne 60.2
(2007), 23552.
12 Helms (n. 11), 129.
13 J. Dunlop, History of Fiction (London, 1814), vol. 1, 59.
14 Cf., e.g., Bowie (n. 4), 4763, at 55.
15 Schmeling (n. 10), 13059 maps out Chaereas characterization against the background of
8

CHAEREAS REVISITED

249

only S. Smiths recent exploration of similarities between Chaereas and Alcibiades has
substantially problematized the widely held idealizing view of Charitons protagonists
in general, and of Chaereas in particular. This article further corroborates such
problematization.
A feature of Chaereas characterization that has triggered disapproval among
students of the genre is his sudden character shift in the seventh book. In the first six
books of the eight-book novel, Chaereas is characterized by passive behaviour that
sharply contrasts him with the resourceful heroine Callirhoe. Unlike her, Chaereas
hardly ever undertakes any action to resolve his problems and spends most of his time
lamenting his separation from his wife. In the seventh book, however, his behaviour
changes dramatically: following the advice of his friend Polycharmus, he joins the
Egyptian army revolting against the Persian king Artaxerxes, and turns out to be a
brilliant soldier. He succeeds in occupying the impregnable city of Tyre, and in less
than no time he is the admiral of the whole Egyptian fleet. Scholars have criticized the
improbability and inconsistency of this character shift. Rohde, for example, articulates the following complaint: man verwundert sich, am Schlu des Ganzen den bis
dahin so wenig energischen Chaereas urpltzlich zum siegreich handelnden und
herrschenden Kriegshelden sich umwandeln zu sehen. Solche Tatkraft stimmt wenig
zu seiner sonstigen Weichlichkeit, zu der Weichlichkeit der ganzen Erzhlung und fast
aller Personen derselben.16
R. Balot argues that Chaereas character shift essentially revolves around the
thematization of martial valour as the fulfillment of the central virtues appropriate
to his gender, training, and elite status.17 In his view, Chaereas military excellence
signposts a turnabout in which Chaereas becomes the man he is required to be if he is
to win back Callirhoe and begin to recreate his marriage (157). D. Scourfield rightly
adds that Chaereas gradually developing ability to learn how to control and to utter
his anger appropriately represents the young mans personal growth towards a full
adult-male status.18 Indeed, Chaereas initial anger, triggered by the false suspicion
about his wifes infidelity, persists throughout the entire novel. What makes Charitons
protagonist an adult man is not a renunciation of anger but the ability to control this
emotion and not be driven to impulsive and irrational behaviour by it. In this paper I
want to draw attention to a new (and yet related) dimension of Chaereas character
shift. I will put forward two arguments. First, I propose that, next to military
excellence and the acquisition of self-control, the acquisition of control over other
characters is an equally important feature of Chaereas character shift.19 In this
connection, I take into account S. Smiths recent, politically oriented reading of
Chaereas as the rising star in the Syracusan political firmament who is about to
traditional concepts of heroism in epic and tragedy (incarnated in Achilles and Ajax respectively). On Chaereas assimilation with epic paradigms, see D. Konstan, La rappresentazione dei
rapporti erotici nel romanzo greco, MD 19 (1987), 927, at 911 and D. Konstan, Sexual
Symmetry. Love in the Ancient Novel and Related Genres (Princeton, 1994), 1617. For a recent
overview of scholarship on Chaereas, see Smith (n. 4), 1922.
16 Rohde (n. 7), 527.
17 R.K. Balot, Foucault, Chariton, and the masculine self, Helios 25.2 (1998), 13962, at 156.
18 D. Scourfield, Anger and gender in Charitons Chaereas and Callirhoe, in S.M. Braund and
G.W. Most (edd.), Ancient Anger. Perspectives from Homer to Galen (Cambridge, 2003), 16384.
19 Smith (n. 4), 834 has concisely touched upon the connection between self-control and
control over other characters as an element underlying the characterization of Charitons two
minor characters Dionysius and Artaxerxes. My reading of Charitons protagonist develops this
point and emphasizes its thematic centrality in the novel.

250

KOEN DE TEMMERMAN

displace Hermocrates soon after his return to Syracuse.20 My paper points out that,
along Chaereas road towards political power, the establishment of control over (and
even manipulation of ) other characters is an issue of primary importance. Secondly,
and consequently, I argue that my observations challenge the widely held view that
character depiction of protagonists in the ancient Greek novel is invariably idealistic.
CHAEREAS AND RHETORICAL CONTROL
The concept of control discussed in this paper is of a rhetorical nature, comprising the
protagonists ability to influence the behaviour of other characters through speech.
In the heyday of the ancient Greek novels, the construction of speech in character
(thopoiia) was one of the so-called progymnasmata, preliminary rhetorical school
exercises in writing and composition. These progymnasmata, discussed by, among
others, Aelius Theon, Ps.-Hermogenes, Aphthonius and Nicolaus,21 were an essential
part of rhetorical education in antiquity from at least the first century B.C. onwards22
and undeniably influenced imperial literature.23 Therefore, it is more than likely that
both writer and reader of narrative will have considered speech an important index of
character. On a more general note, M. Gleason has extensively discussed rhetorical
performance as a crucial tool in achieving and displaying manliness in the first
centuries A.D.24 Within this framework, this article sets out to interpret Chaereas
rhetorical performance and self-presentation as an index of his growth towards male
adulthood.
The distribution of Chaereas speeches, including public speeches and private
conversations, around his character shift is significant. In the first six books (before
the character shift, that is), Chaereas speaks in public only twice. The last two books
feature no less than seven such speeches. Let me first discuss the speeches before his
or self-accusation
character shift. Chaereas first public speech is his
after the supposed murder of Callirhoe in Book 1 (1.5.45). The primary narrator
highlights that Chaereas adduces none of the arguments in his defence. Instead, he
asks the jury to be sentenced to death for murdering the daughter of Syracuses first
citizen Hermocrates. Moreover, he insists on being denied burial after his death,
comparing his crime to temple robbery and parricide. S. Smith rightly argues that the
equation between Chaereas emotional expression and the sincerity of his inner state
20 Smith (n. 4), 1901. See also S. Lalanne, Une ducation grecque: rites de passage et construction des genres dans le roman grec ancien (Paris, 2006), 158 for a short depiction of Chaereas
as the new Hermocrates.
21 All treatises are edited by L. Spengel (ed.), Rhetores Graeci (Leipzig, 1854 and 1856), vols 2
and 3. A more recent edition of Theon is M. Patillon (ed.), Aelius Thon. Progymnasmata (Paris,
1997).
22 Cf. G. Kennedy, Classical Rhetoric and its Christian and Secular Tradition from Ancient to
Modern Times (Chapel Hill / London, 1999), 27.
23 Cf. A. Cizek, Imitatio et tractatio. Die literarisch-rhetorischen Grundlagen der Nachahmung
in Antike und Mittelalter (Tbingen, 1994), 23641; H. Cichocka, Progymnasma as a literary
form, SIFC 10 (1992), 9911000; G. Kennedy, Greek Rhetoric under Christian Emperors
(Princeton, 1983), 53 and 143. Theon, Prog. 70.2430 Spengel 2 explicitly underlines the importance of progymnasmata in contemporary literature:

24 M. Gleason, Making Men. Sophists and Self-Presentation in Ancient Rome (Princeton,

1995), xxxxix and 13168.

CHAEREAS REVISITED

251

in this speech signposts neutralization of the power of rhetoric.25 This neutralization

is, I think, equally fleshed out by the fact that Chaereas speech in the end does not
generate the envisaged effect: the audience forgets about the dead generals daughter
and starts to commiserate with Chaereas (
, 1.5.6). When the jury finally acquits Chaereas, the protagonist himself,
unhappy with this decision, thinks of possible ways to kill himself (
). Smith (1267) acknowledges that the
power of rhetoric, absent from this speech, is amply demonstrated elsewhere in
Charitons novel. He specifically refers to other male characters, such as Dionysius
and Artaxerxes, who, unlike Chaereas in this episode, rationalize personal emotions
by means of subtle, rhetorical self-fashioning. As I will point out, Chaereas own
rhetorical self-fashioning later in the story will also demonstrate the importance of
the power of rhetoric in this novel.
Chaereas second public speech is found in Book 3 (3.4.56 and 3.4.15), when he
returns from the search for Callirhoe in the Ionian Sea and brings Theron to Syracuse.
Framing this speech, the primary narrator takes pains to indicate that Chaereas has
trouble assuming the self-control that might be expected of an orator addressing his
audience. I refer to the physical details preceding the quotation of his speech (
, 3.4.4), the uncontrolled outbursts of emotion (
) and
the inability to address the public verbally (
, 3.4.4) or
visually (
, 3.4.5). The first words of the speech itself, moreover,
are in line with the overall picture drawn by the narrator. Chaereas says that it is not
the right time for rhetoric, but for mourning (
, 3.4.5). As in the first speech, the power of rhetoric is neutralized, this
time explicitly. Again, this neutralization is signposted by Chaereas failure to achieve
the aim envisaged by his public performance. After Therons confession about
Callirhoes abduction, Chaereas asks that Therons life should be spared in order to
facilitate the search for Callirhoe (3.4.15). This request, however, is denied. The
, 3.4.18)
assembly follows Hermocrates counterarguments (
and Theron is executed.
In neither of the two speeches, then, is Chaereas capable of persuading his
audience to grant his requests. A comparable lack of control over his narratee(s)
characterizes his private speeches in the first half of the story. Moreover, in these
dialogues, it is Chaereas himself who is systematically controlled by his interlocutors.
In fact, many of his private speeches and dialogues suggest that Chaereas is easily
misled, a characteristic explicitly attributed to him by the primary narrator in a
, 7.1.4). The very first
gnom or maxim (
words uttered by Chaereas in the novel are emblematic of this characteristic. They
form only one sentence and reproach Callirhoe for being responsible for traces of

partying outside the newly married couples house:

(It is what has happened to me that I am crying about; you
have forgotten me straightaway!, 1.3.5). These words are set in a highly emotional
context, involving Chaereas anger and grief (

, 1.3.45),
emphasized by the primary narrators heavily elaborated account of Chaereas
, 1.3.4;
body language (
, 1.3.5). Unlike the protagonist himself, however,
the reader has been informed by the primary narrator that Chaereas is being misled by
25

Smith (n. 4), 126.

252

KOEN DE TEMMERMAN

the suitors, who left evidence of a party at the house the night before in order to
trigger Chaereas anger and suspicion of his wife. In these circumstances, Chaereas
(I cry) will turn out to be emblematic of his behaviour when
very first word,
confronted with misfortunes in the following six books of the novel.26
Chaereas characterization as an object of deception by other characters is echoed
in his depiction as an internal narratee in the first six books of the novel. From the
very beginning of the novel, Chaereas is controlled and manipulated by other
characters speeches. A first example is the dialogue between Chaereas and the tyrant
of Acragas accomplice, who fools Chaereas into believing that his wife Callirhoe has
been unfaithful (1.4.78). Chaereas is devastated by the news and asks his interlocutor
to witness the adultery with his own eyes. This request plays, of course, precisely into
the hands of the conspirators, and allows the interlocutor to set up the meeting
between Chaereas and Callirhoes alleged adulterer.
My second example of Chaereas as dupe is the conversation between Chaereas
and Mithridates (4.4.25). This dialogue occurs when Chaereas has just been
informed by Mithridates that Callirhoe has married Dionysius in Miletus. Chaereas
asks Mithridates permission to go to Miletus and claim his wife from Dionysius.
Mithridates, however, advises against this plan and suggests that Chaereas write a
letter to Callirhoe first. He prefaces his advice as follows:

(4.4.2)
As far as I am concerned ... you can go. I dont want you to be separated from your wife even for
one day. I wish that you had never left Sicily and that no trouble had ever befallen the two of
you.27

The reader knows that Mithridates words aim at fooling Chaereas into believing that
he is truly concerned about Chaereas love for Callirhoe. Unlike Chaereas, the reader
has been informed shortly beforehand about Mithridates hope that, while Dionysius
and Chaereas quarrel about Callirhoe, he himself will be able to become her lover
(4.4.1). The primary narrator explicitly clarifies this strategy when explaining why
Mithridates rejoices in Chaereas sad story: Chaereas grief gives him the opportunity
to talk and take action about Callirhoe in order that he would appear to be helping
, 4.3.11). The reader, who has been repeatedly
a friend (
informed about Mithridates love for Callirhoe (4.1.9, 4.2.4), realizes that Mithridates
tries to profit from the situation at Chaereas expense. Chaereas, on the other hand,
has no idea about his hosts plan and thinks that he truly wants to help him. In his
,
letter to Callirhoe, he even calls Mithridates his benefactor (
4.4.7). At the end of the story, when narrating his adventures in front of the
Syracusan people, he equally characterizes Mithridates as a true helper (
, 8.8.4).
All the above speeches, both public and private, characterize Chaereas in a similar
way. In public speech he is not able to persuade his audience to approve his requests.
In private conversation his lack of control is highlighted by the control exerted upon
26 Cf., e.g., Chariton, Callirhoe 1.6.5, 3.3.14, 3.4.4, 3.6.6, 4.4.6, 5.2.4. The contrast between the
resourceful Greek novel heroines and their helpless male counterparts in general has been
addressed by, among others, Konstan (n. 15 [1994]), 1526.
27 English translations of Charitons text are taken from B.P. Reardon (ed.), Collected Ancient
Greek Novels (Berkeley/Los Angeles/London, 1989) and slightly modified where necessary.

CHAEREAS REVISITED

253

him by other characters. Let us now turn towards Chaereas speeches after his
character shift in the seventh book. His first words after this shift are emblematic of a
new strand in his characterization. This speech, addressed to the leader of the
Egyptian rebels, occurs when Chaereas and Polycharmus have been taken prisoner by

, 7.2.2). When they

the Egyptians, who regard them as spies (
are brought before the leader, Chaereas, without waiting to be addressed, embarks on
a speech in which he reminds the audience of his homeland and noble descent (
, 7.2.3). Moreover, he associates
himself with the famous Hermocrates by mentioning his marriage to his daughter
Callirhoe (7.2.3). S. Smith correctly identifies this speech as the first instance of
Chaereas participation in the same rhetoric of self-representation adopted by other
Syracusans in the story.28 Chaereas relies on his ties to Hermocrates to secure a safe
entrance into the Egyptian army. Ultimately, he expresses his and his friends desire to
die fighting against the Persians. Chaereas attention to self-depiction is easily identifiable as a classical rhetorical technique, well documented by Aristotles famous
account of the importance of favourable character construction (thos) as one of the
tools ensuring the audiences persuasion.29 Significantly, Chaereas rhetorical strategy
is successful: the Egyptian leader welcomes them into the army and provides them
with arms and a tent. In this speech, Chaereas exploits his own origins in order to
achieve a specific aim and, simultaneously, takes rhetorical control over his narratee
for the first time. By creating a favourable characterization of himself and his friend,
he succeeds in becoming a soldier in the Egyptian army, thus effecting the plan
suggested by Polycharmus earlier (7.1.11).
A study of Chaereas remaining public speeches after his character shift corroborates and, simultaneously, develops this point about the protagonists changing
rhetorical abilities. In the following overview, I focus on a number of rhetorical
techniques adopted by Chaereas. These observations will be contextualized by an
account of how the primary narrator enhances his protagonists characterization in
these passages.
In a number of speeches Chaereas characterization as a rhetorically successful
soldier and general is constructed, both by the primary narrator and by Chaereas
himself, through assimilation with mythological and historical paradigms. Chaereas
speech in the assembly of the Egyptian army generals (7.3.45) is a case in point.
Since the impregnable city of Tyre is an obstacle to the Egyptian military advance,
their leader proposes to retreat. At this proposal all are silent and downcast (
, 7.3.3), except Chaereas (
). The Egyptian leaders speech and Chaereas reaction to it are
modelled on an episode in the ninth book of the Iliad (9.1728 and 9.3249), where
Agamemnons proposal to return to Greece is countered by Diomedes. This parallel is
suggested by a number of elements. Like the Egyptian leader, Agamemnon apostro, Il. 9.17;
phizes the members of the assembly as friends (

, 7.3.2) and proposes returning home. Secondly, like the Egyptian

leaders speech, Agamemnons speech triggers silence (

, Il. 9.29) and

sadness (
, Il. 9.30) in the audience. The reaction of the audience to
Diomedes speech, thirdly, coincides with the assemblys reaction to Chaereas speech:
in the Iliadic episode, all listeners loudly applaud Diomedes speech (

, Il. 9.501); in Chariton, likewise, all listeners are too ashamed not to
28

Smith (n. 4), 88.

29

Arist. Rh. 1356a515.

254

KOEN DE TEMMERMAN

approve Chaereas proposal (

, 7.3.6). Fourthly, and finally, Chaereas answer itself unmistakably evokes
Diomedes:

(7.3.5)

But if you insist on going, leave a few volunteers with me; I and Polycharmus will fight, for it is
at a gods behest that we have come.

These words are an adaptation of the conclusion to Diomedes speech (Il. 9.489).
They evoke Diomedes forecast about Agamemnons plan to abandon the war: if
Agamemnon wants to flee, the Greeks will remain. And if they want to flee as well, at
least Diomedes himself and his companion Sthenelus will remain.30 It has often been
noted that both explicit and implicit assimilation of Chaereas with epic heroes is
frequent in Chariton.31 What is important in this passage, however, is that it is not
merely the primary narrator who casts Chaereas as an epic hero, but also Chaereas
himself, adopting the above-mentioned Homeric quotation in his own speech.32
Chaereas presents himself as an epic hero and soldier. This strategy is successful
and Chaereas achieves his aim: the Egyptian leader abandons his plan to retreat and
gives Chaereas as many soldiers as he wants to capture Tyre (
).
A similar pattern appears almost immediately afterwards, when Chaereas
addresses 300 Dorian soldiers (

, 7.3.7) whom he has chosen to

capture Tyre (7.3.810). In this speech, which is an adaptation of the speech delivered
by Xenophon to his men before they engage in battle against their Persian enemy
(Xen. An. 3.2.732),33 Chaereas is once again assimilated with literary paradigms.
Again, Chaereas assimilates himself with two famous heroes, Leonidas and
Othryades, both, like Chaereas, leaders of 300 Spartans/Dorians at Thermopylae and

at Thyrea respectively. This time, the association is explicit (

, 7.3.11) and echoed by the result of Chaereas speech, which is that his
soldiers declare him their leader. Apart from assimilating himself with historical
leaders, Chaereas adopts a number of other rhetorical techniques in order to pave the
way for this decision. He starts, for example, by identifying his audience as the best
, 7.3.8). Subsequently, he aligns
men in the army (
himself with his men by referring to their common virtues of
and
, and
by contrasting himself and his audience with their Tyrian enemies. Finally, he does
not propose that he himself should take command, but declares himself willing to
serve under one of the 300 Greek soldiers.

30

Hom. Il. 9.423 (

,|
).
Cf., among others, M. Biraud, Lhypotexte homrique et les rles amoureux de Callirho
dans le roman de Chariton, in A. Goursonnet (ed.), Smiologie de lamour dans les civilisations
mditerranennes (Paris, 1985), 217; P. Robiano, La citation potique dans le roman rotique
grec, REA 102 (2000), 50929; E. Cueva, The Myths of Fiction. Studies in the Canonical Greek
Novels (Ann Arbor, 2004), 245.
32 On characters assimilation of themselves with mythological paradigms as a rhetorical
device, cf. also Smith (n. 4), 104.
33 For details, cf. Smith (n. 4), 1725.
31

CHAEREAS REVISITED

255

As well as assimilating himself with historical and mythological paradigms,

Chaereas adopts other rhetorical techniques to persuade his audience. When, now as
the admiral of the entire Egyptian fleet, he arrives with his army on Cyprus, he
informs some of his troops that their Persian enemies have beaten the Egyptian land
forces and killed their leader (8.2.1011). The success of this speech, which is to
persuade the soldiers to abandon the war and return home, is achieved mainly by two
rhetorical techniques. First, the primary narrator emphasizes that Chaereas selects his
internal narratees. His speech is not directed towards the entire army, but to his
captains, the 300 Greek soldiers and all the Egyptians whom he saw to be well
, 8.2.9). Secondly, Chaereas subtly controls
disposed to him (
and manipulates his audiences decision-making process. Before depicting their hopeless military position and addressing the fact that they are surrounded by enemies
(
, 8.2.10), he reminds his soldiers of
the importance of unity in their naval military successes up to that point (
, 8.2.10).34 Subsequently, he suggests capitulation to the
Persian king as the only possible solution. His audiences refusal to agree does not
come as a surprise, either to the reader or to Chaereas himself. At last one of the
soldiers called Brasidas proposes to return to Sicily. While everyone applaudes this
suggestion (
, 8.2.13), Chaereas pretends to disapprove
(
), adducing the length of the
journey as a pretext (
) for his scepticism. The primary narrator, for
his part, informs the reader that Chaereas only wants to test the audiences firmness
) by pretending to
of purpose (
disagree. When the troops insist, Chaereas is persuaded to go home. Thus, Chaereas
reaches a consensus about terminating the war and returning home, without proposing this solution himself.
This episode is replete with Homeric resonances relevant to Chaereas characterization. Chaereas reaction to Brasidas proposal clearly echoes Agamemnons
attempt to manipulate the army in Iliad 2.53154.35 In this passage, Agamemnon
proposes withdrawal from the war and urges the soldiers to return home (
, Il. 2.140). As announced by the primary narrator
, Il. 2.55) and by Agamemnon himself in a speech directed only to
(
the members of the council (Il. 2.705), this is a ruse, ultimately intended to make the
soldiers more eager to participate in a planned attack on Troy (
, Il. 2.72). Both in Homer and Chariton, then, the generals
attempt to manipulate the army is designed to test the soldiers (
, Il. 2.73;
, Chariton 8.2.13). The crucial difference between the two episodes is,
of course, that Agamemnons stratagem fails and results in chaos: the soldiers
immediately run to the ships to prepare for departure. Significantly, order is not
restored until Odysseus rhetorical skills in a lordly manner brought the army under
, Il. 2.207). Unlike Agamemnon, Chaereas
control (
does not propose returning home, but subtly paves the way for this suggestion by
proposing capitulation to the Persian king, well knowing that his audience will not
agree. In both cases, the audiences agree with the proposal to return home, but
whereas Agamemnon expects the opposite reaction from his troops, Chaereas subtle
34

Cf. Smith (n. 4), 97.

The implicit presence of this Iliadic episode has been touched upon by G.P. Goold,
Chariton. Callirhoe (Cambridge, MA/London, 1995), 375, and Smith (n. 4), 97, but has not been
dealt with in any detail.
35

256

KOEN DE TEMMERMAN

demagoguery aims at triggering exactly this response.36 The Iliadic subtext in this
episode, therefore, depicts Chaereas as a non-Agamemnon: whereas Agamemnons
testing of the soldiers unexpectedly results in chaos, Chaereas testing of the soldiers
triggers a confirmation of their resolution to go home, as anticipated by Chaereas.
Whereas Agamemnon eventually needs Odysseus rhetorical skills to restore order
after his own unsuccessful rhetorical performance, Chaereas becomes an Odysseus
himself in successfully manipulating his audience.
The episode involving Agamemnons unsuccessful attempt to arm the soldiers for
battle is not the only subtext underlying Chaereas speech in Book 8. The context in
which the speech is set equally evokes the above-mentioned discussion between the
Egyptian leader and Chaereas in the seventh book. This parallel is highlighted by
the narrators twofold explicit statement that Chaereas reaction to the proposed
,
solution to the crisis is different from the rest of the audiences (
7.3.3;
, 8.2.13). More importantly, however, the assimilation of the
latter episode with the former reactivates the Iliadic episode of the disagreement
between Agamemnon and Diomedes (Il. 9). In all three episodes, a military
commander accentuates the armys hopeless position, after which the possibility is
raised of abandoning the war and going home. Some verbal echoes underline the
association between all three episodes. First, Chaereas opening words echo the
Egyptian leaders and Agamemnons apostrophes (

,
7.3.2;
, 8.2.10;
, Il. 9.17). Second, the
audiences reaction to Chaereas speech about their hopeless position corresponds
with the audiences reaction to the Egyptian commanders and Agamemnons
, 7.3.3;
, 8.2.12;

speeches: in all three cases, it is silence (

, Il. 9.29). Third, Brasidas speech, offering the solution to the problem that will
, 8.2.13),
eventually be chosen, is noted as being applauded by all listeners (
which echoes the reception of Chaereas and Diomedes speeches in the two mirror
, 7.3.6;
scenes (

, Il. 9.501). The crucial difference between the two episodes

in Chariton, however, is that, whereas the first episode associates Chaereas with
Diomedes, the second portrays Chaereas himself as the leader of the army who
informs the troops about their hopeless military position. This role aligns him, of
course, with Agamemnon. Moreover, the result of Chaereas speech is, in the end,
precisely the aim envisaged also by Agamemnon in Book 9, namely to abandon the
war and return home. Significantly, the point of the association lies in the obvious
difference between the two heroes: whereas Agamemnons proposal is criticized and in
the end rejected, Chaereas manages to achieve his aim. More interestingly, he does so
without proposing this solution but by creating the illusion that he himself is being
persuaded by a suggestion from the audience. Thus, he is able to engineer withdrawal
from the war without running the risk of being characterized by his troops as a fool or
a coward, two characteristics attributed to Agamemnon by Diomedes because of his
proposal to withdraw.37 Like the echoes from Iliad 2, the echoes from Iliad 9 also
depict Chaereas as a non-Agamemnon. Unlike Agamemnon, Chaereas controls his
listeners by giving them the impression that they are in control themselves. Whereas
36 On Chaereas characterization as a demagogue in this passage, cf. also Smith (n. 4), 98, who
describes Chaereas as a kind of Cleon or Alcibiades, using subtle rhetorical persuasion as a
means of demagoguery.
37 Diomedes reproaches Agamemnon for being a fool (
, Hom. Il. 9.32) and
displaying a lack of courage (
, Il. 9.39).

CHAEREAS REVISITED

257

Chaereas himself was controlled by interlocutors in the first half of the story, he has
become a public speaker whose rhetorical qualities surpass those of one of his most
important epic paradigms.
A fourth rhetorical technique adopted by Chaereas to control his audience is
distortion of the truth. In his speech addressed to the Tyrians (7.4.5), he uses a lie
which eventually leads to the capture of Tyre. He and his soldiers approach the city
gates and tell the Tyrians that they are mercenaries deserting the Egyptian army. The
stratagem is successful: after the Tyrians have opened the gates, Chaereas and his
troops take the allegedly impregnable city. It is worthwhile pointing out that earlier in
the story, the protagonist uses a similar lie to gain access to the ranks of the Persian
army (7.2.1). In order to cross the Euphrates in the Persians wake, he and
Polycharmus had claimed that they wanted to join the army. Unlike the Persians, the
reader knew that this was a lie, since the two friends wanted to cross the river to join
not the Persian but the Egyptian army. From the capture of Tyre onwards, the tactful
use of lies and incorrect information will increasingly become part of the protagonists rhetorical strategy. Chaereas public speech in 8.2.5 provides an excellent
example. It occurs when Chaereas, who has become admiral of the entire Egyptian
fleet, has been informed by a messenger of the defeat of the Egyptian land forces and
the death of their leader. The messenger states, moreover, that the Persian enemy is on
its way to Aradus, the island harbouring Chaereas fleet. The primary narrator
explicitly refers to Chaereas subsequent speech to the Egyptian sailors as a ruse or
(8.2.5): Chaereas tells the sailors that the Egyptian army has defeated the
Persians, and orders them to set sail without specifying their destination. Interestingly, Chaereas use of this scheme results from Callirhoes intervention. When
Chaereas leaps up after hearing the bad news, Callirhoe advises him not to make it
public (8.2.4). She argues that this would cause revolt among the troops and that we
shall be captured again and shall be worse off than ever (
, 8.2.5). The primary narrator comments that
Chaereas is soon convinced by this advice (
, 8.2.5). In
Callirhoes words, the issue of controlling and being controlled is explicitly
highlighted as the main reason why Chaereas should not give an accurate account of
what has happened. After his speech, as all the sailors are preparing to depart,
Chaereas takes advantage of the confusion in the harbour (
, 8.2.7) to order his captains to set sail for Cyprus secretly. Once
they arrive at Paphos the next day, they are safe from immediate danger. Thus
Chaereas public address to the naval troops clearly generates the desired effect:
thanks to the
, Chaereas restrains his troops from mutiny and manages to keep
them under control. Callirhoe has taught him that control over others can be achieved
by rhetorical devices involving distortion of the truth.
Chaereas last public speech (8.7.98.11) thematizes some important issues of
manipulation and distortion of the truth already present in earlier speeches. Significantly this speech, in which Chaereas reports his adventures to the Syracusan people
upon his homecoming, constitutes Chaereas new identity before his fellow Syracusans.38 The narrator gives some important background information about this speech.
First, it is not Chaereas who insists on telling the story, but the Syracusan crowd who
insist on hearing it, after having led him to the theatre (
, 8.7.3). Second,
38

Cf. Smith (n. 4), 142.

258

KOEN DE TEMMERMAN

Chaereas starts with the last events of his story (

, 8.7.3) because
he did not want to cause the people sorrow by telling them of the grim episodes at the
beginning. When the crowd protests, however, insisting that he does not omit
, 8.7.3),
anything (
Chaereas hesitates because he is ashamed to talk about many events that had not
turned out to his satisfaction (
, 8.7.4). The combination of the audiences explicit
request not to omit anything and the narrators equally explicit statement about
Chaereas hesitation must surely alert the reader as to whether the information
provided in Chaereas speech is actually complete.39 S. Smith, moreover, has recently
drawn attention to various sorts of deceptions and conceits in the scene when
Chaereas returns to Syracuse. This scene, reminiscent of Alcibiades triumphant
return to Athens after exile, is filled with elements that consistently trigger incorrect
inferences from the Syracusan people about what is happening.40 The issue of
deception activated by these elements provides the framework in which Chaereas
speech should be read. I argue that Chaereas account of his adventures is in some
instances manipulatory and deceptive, diverging significantly from the primary
narrators account in the foregoing chapters of the novel.
The first relevant passage in Chaereas speech is his account of his and
Polycharmus discovery of Callirhoes statue in the temple upon their arrival in
Miletus:

(8.8.1)

At the time, when I had landed on this estate, I saw only Callirhoes statue in a temple, and that
gave me great confidence. But during the night a band of Phrygian brigands made a lightning
raid on the shore, set fire to our ship

Chaereas contrasts the unfortunate outcome of the episode, due to the brigands
attack, with his own confidence in a good outcome after having seen Callirhoes
statue. The reader, however, recalls that Chaereas did not have great confidence in
this episode. In fact Chaereas fainted when he saw the statue of his wife, a reaction
emphatically marked with a Homeric quotation by the primary narrator (3.6.3). The
temple servant even had to bring water to resuscitate him. Moreover, the primary
narrator emphasized Chaereas lack of self-control in this episode by contrasting the
,
protagonist with his friend Polycharmus, who was able to control himself (
3.6.5) and prevented Chaereas from betraying who they were (
, 3.6.5). Furthermore, when he was alone, Chaereas
threw himself on the floor and deplored his situation in a lamenting monologue. In
short, the reader recalling this episode while reading Chaereas report realizes that
great confidence is not a correct representation of what had happened. Chaereas,
however, understandably chooses to omit his lack of self-control in his version of the
story.
As to why Callirhoe married Dionysius, Chaereas says the following:
39 On checking a characters direct speech against the primary narrators account, cf. T. Hgg,
Narrative Technique in Ancient Greek Romances. Studies of Chariton, Xenophon Ephesius, and
Achilles Tatius (Stockholm, 1971), 253.
40 Smith (n. 4), 231.

CHAEREAS REVISITED

259
(8.7.11)

When Callirhoe realized that she was pregnant by me, she found herself compelled to marry
Dionysius, because she wanted to preserve your fellow citizen.

That Callirhoe is compelled to marry Dionysius, is, indeed, confirmed by the primary
(5.1.1) to refer to Callirhoes marriage.
narrator, who likewise adopts the term
Chaereas is equally correct in adducing his child as the reason for Callirhoes decision
to marry. The antonomasia used to refer to the child, however, is significant. By calling
, Chaereas seems to be suggesting that Callirhoes loyalty
his child
towards her home city played a role in this decision,41 which was not the case
according to the primary narrators version. Chaereas thus colours his story in order
to generate the audiences sympathy for his wife.
Chaereas desire to generate sympathy for his wife might be responsible for the
distortion of some other details of the true story. In his account of his imprisonment on Mithridates estate, he states that Mithridates discovered his identity after
Polycharmus uttered his name under torture (
, 8.8.3). The reader,
however, recalls that Polycharmus did not utter Chaereas name, but Callirhoes and
that he did so not under torture, but when Polycharmus and Chaereas were carrying
their crosses to the execution site. Moreover, Polycharmus utterance was a fierce
reproach:

(4.2.7)

Polycharmus, as he carried his cross, said: Callirhoe, it is because of you that we are suffering
like this! You are the cause of all our troubles!

To mention this detail in front of the entire Syracusan people would be embarrassing
both for Chaereas friend Polycharmus and for his wife Callirhoe. It is therefore
reasonable to assume that Chaereas decides to omit this detail in the official version.
Chaereas account of the events is also characterized by a tendency to emphasize
the heros own achievements at the expense of the achievements of other characters.
He recounts his arrival amongst the Egyptian army, for example, as follows:

(8.8.8)
The Queen took Callirhoe with her, and I heard a false report someone told me she had been
awarded to Dionysius. To get my revenge on the King I went over to the Egyptians and brought
off great feats: by my own actions I subdued Tyre, which was very difficult to take; I was then
appointed admiral, beat the Great King at sea, and captured Aradus, where the King had left the
Queen for safety, along with the riches you have seen.

The emphasis laid by Chaereas upon his own achievements is significant. Accordingly, he completely omits the role played by Polycharmus in this important episode.
As the reader recalls, Chaereas burst into a lament and wanted to commit suicide after
hearing the false report of Callirhoe that he presents here as the starting point of his
41

See also Smith (n. 4), 223.

260

KOEN DE TEMMERMAN

personal aristeia (7.1.47). It was Polycharmus who came up with the idea of harming
their enemy with their own death. Again, Chaereas subtly adapts an episode which
reveals his embarrassing lack of self-control. Instead of telling the truth, he credits
himself with the decision to join the Egyptian army and immediately proceeds with
.
recounting his own
This pattern is repeated in the protagonists account of how he managed to secure
the Persian kings friendship for the Syracusan people:

(8.8.10)
Finally, I secured the Great Kings friendship for you by making a present to him of his wife and
by sending to the Persian nobles their mothers and sisters and wives and daughters.

At this point, the reader vividly recalls that it was Callirhoes idea, not Chaereas, to
release the Persian queen (8.3.1). Significantly, Chaereas had been blushing while
admitting to his wife that he wanted to take the queen to Syracuse as a slave (8.3.1).42
Moreover, in his letter to Artaxerxes, Chaereas did admit that it was not his but
Callirhoes idea to release Statira (

, 8.4.3). Whereas
Chaereas presents the kings friendship for Syracuse as his personal achievement, the
reader realizes that there would not be any such friendship if Callirhoe had not
intervened. In fact, the idea of keeping the Persian queen as a prisoner had explicitly
, 8.3.2). Again, Chaereas credits
been referred to as madness by Callirhoe (
himself with someone elses achievements.
This analysis explains why Chaereas at first is not willing to recount his adventures.
The story contains a number of episodes about which he should rightly be ashamed.
It is significant that, by the end of the novel, Chaereas is capable of distorting and
covering up these episodes. It is equally worthwhile to note that, after his speech, his
request that his sister be given in marriage to Polycharmus is accepted. Unlike in his
two speeches to the Syracusans at the beginning of the story, Chaereas has become an
orator who is able to control his audience.
CONCLUSION
The above observations about Chaereas changing rhetorical ability lead me to
suggest that the strand of self-control in the protagonists characterization acknowledged by D. Scourfield43 can be complemented by the acknowledgment of an equally
important strand of rhetorical control over other people. In addition to the transition
from lacking self-control to adopting self-control in mastering anger, Chaereas
character displays a significant transition on the level of rhetorical performance. At
the beginning of the story, Chaereas is unable to achieve his desired aims through the
use of speech. He lacks the rhetorical control required to persuade his audience.
Moreover, in private conversation, he is controlled, and even manipulated and
deceived, by his interlocutors. From the seventh book onwards, however, he develops
the rhetorical ability to persuade his audience through the manipulation of speech.
This ability is reflected in various features. First, he is successful in constructing
favourable characterizations of himself to ensure his audiences persuasion, a
42 On the significance of blushing in Chariton (and this scene in particular), see De
Temmerman (n. 11), 23552, at 2478.

CHAEREAS REVISITED

261

traditional rhetorical device discussed by Aristotle as the construction of thos.

Second, Chaereas associates himself, explicitly and implicitly, with mythological
(epic) and historical heroes to enhance this construction (e.g. 7.3.45 and 7.3.810).
Third, he is aware of the importance of consciously selecting his internal narratees
(e.g. 8.2.1011). A fourth technique consists in subtly manipulating his audience by
guiding his listeners towards a specific decision while giving the impression that they
have freely and independently reached it (e.g. 7.3.810 and 8.2.1011). Fifth and
finally, Chaereas realizes the importance of distorting the truth to achieve control
over his audience. The assumption that rhetorical control involves conveying
information that does not necessarily correspond to reality underlies all his public
speeches from the capture of Tyre onwards. It is this last technique in particular that
culminates in his last public speech addressed to the Syracusan people. His account of
his and Callirhoes adventures is greatly concerned to cover up embarrassing details
of the story. By omitting details compromising his own behaviour and by crediting
himself with other persons achievements, Chaereas characterizes himself more
favourably than the primary narrator does in the foregoing story.
Why does the narrator depict this development in his heros rhetorical abilities in
the later books of the novel? First, my observations are in line with, and offer an
interesting addition to, S. Smiths recently developed argument that Charitons novel
implicitly tells a story about the transition of political leadership from Hermocrates
to Chaereas.44 Within such a transition, the achievement of rhetorical control is, as I
have pointed out, of crucial importance. From a broader perspective, my reading of
Chaereas deepens S. Lalannes thesis that the ancient Greek novels embody the
protagonists rite of passage from childhood to mature adulthood.45 In her view, the
heroes and heroines many ordeals and adventures function as preparations for their
tasks as socially accepted citizens and wives respectively.46 For male characters, this
paideia is primarily directed towards the acquisition of a number of basic qualities
such as moderation, perseverance and magnanimity, which are emblematic of the
virtues of a civilized Greek male adult. In the case of Chaereas I argue that Chariton
thematizes the importance of rhetorical skilfulness as yet another essential quality of
male adulthood.47 In addition to the military achievements marking Chaereas entry
into manhood from Book 7 onwards, his ability to perform successfully on the battlefield of rhetoric is at least as important. It is no coincidence, therefore, that Chaereas
first active resistance against the misfortunes befalling him occurs in the Babylonian
courtroom during the trial regarding the validity of his and Dionysius claims to
Callirhoe. Although his intervention in this case is limited to the interjection of brief
reproaches to Dionysius and a number of arguments corroborating his claim (5.8.5),
its rhetorical setting is proleptic of the important place that rhetoric will occupy in
Chaereas life once he decides to take control of his own destiny. Moreover, Charitons
language assimilates this trial with warfare (
),48 thus implicitly marking Chaereas first attempt to intervene actively in

43

Scourfield (n. 18), 16384, at 16375.

Smith (n. 4), 1901.
45 Lalanne (n. 20).
46 See Lalanne (n. 20), 16.
47
See also R. Webb, Rhetoric and the novel: sex, lies and sophistic, in I. Worthington (ed.), A
Companion to Greek Rhetoric (Malden, MA/Oxford/Carlton, 2007), 52641, at 534, who touches
upon the link between rhetoric and masculinity in Chaereas last speech in Syracuse (8.7.98.11).
48
Lalanne (n. 20), 16.
44

262

KOEN DE TEMMERMAN

the course of events as an emblem of two major areas in which he will excel in the
succeeding books.
The above discussion provides a dimension to the characterization of Charitons
male protagonist that has remained relatively unexplored so far. First, Chaereas
characterization cannot be adequately described by merely addressing the notion of
typical character. Rather, it requires attention to specific aspects of individualization.
Second, his characterization does not merely thematize the oft-noted transition from
helplessness and weakness to courage and strength, but seems to suggest that, like
self-control, the ability to control other people by the power of rhetoric is an essential
tool to be acquired on the road towards male adulthood. This suggests that Charitons
male protagonist does not fit into the clear-cut and somewhat monolithic view
developed recently on male protagonists in the novels as lacking rhetorical skills
altogether.49 Third, I think that Chaereas characterization has a much more realistic
dimension than has been identified by existing scholarship. In Chariton, becoming an
adult male citizen involves developing awareness of the importance of rhetorical
control, manipulation and deception, all abilities that display a much closer relationship to psychological realism than to idealism. The widely held view that Charitons
Chaereas is the prototype of the ideal novel hero should therefore be revised.
Instead, it seems to me more plausible that Chariton consciously develops a critical
stance towards idealistic character depiction in the novelistic genre. Therefore,
ultimately, I do not believe that the widely adopted classification of ancient novelistic
texts into ideal and realistic texts allows us to capture the peculiar position of
Charitons novel.50
Ghent University

KOEN DE TEMMERMAN
koen.detemmerman@ugent.be

49 K. Haynes, Power of the prude: configurations of the feminine in the Greek novel, Ancient
Narrative 1 (2001), 7392, for example, highlights novelistic heroines rhetorical qualities in
contrast with the lack of such qualities in their male counterparts. In Fashioning the Feminine in
the Greek Novel (London, 2003), she develops this point at greater length.
50 An earlier version of this paper was presented in a Kyknos panel on the ancient novel at the
Celtic Conference in Classics at the University of Wales, Lampeter (AugustSeptember 2006). I
would like to thank Anton Powell for the conference organization, John Morgan and Meriel
Jones for the organization of the panel, the audience for their comments, and Kristoffel Demoen,
Graeme Miles, Susan Stephens and the anonymous referee of CQ for suggestions on written
versions. Any errors or oversights are entirely my own.

Classical Quarterly 59.1 263269 (2009) Printed in Great Britain

doi:10.1017/S00098388090000202

263
A PINDARIC
CHARIOTEER
JANET
DOWNIE

A PINDARIC CHARIOTEER: AELIUS ARISTIDES

AND HIS DIVINE LITERARY EDITOR
(ORATION 50.45)
In his fourth Hieros Logos (Or. 50), Aelius Aristides reports that Asclepius instructed
him in a dream to dedicate a victory tripod commemorating his recent choral performances. Aristides composes an epigram to be inscribed on the monument, detailing
his efforts as poet, president of the contests, and chorgos:
/

(Or.
50.45).1 But the god rejects this poetic attempt, dictating to his patient instead verses
that celebrate Aristides as an illustrious charioteer of ever-flowing tales:
/
(Or.
50.45). Although no archaeological trace of the memorial has come to light at the
Asclepieum in Pergamum, the story has attracted some interest as indirect evidence
for dedicatory practices of the sanctuarys elite visitors, and scholars have added the
revised epigram to the repertoire of inscriptions that illustrate how orators of the
imperial era publicized their status.2 Little attention has been paid, however, to the
nature of Aristides self-commemoration in its narrative context.3 The first dedicatory
epigram is Aristides own composition; the revised version, on the other hand, is the
result of divine intervention, and it incorporates a striking metaphor of the poet as
charioteer. I shall argue that Aristides borrows this epinician motif from one of the
classical authors he most admired Pindar in order to make a statement about his
own relationship to his patron god, Asclepius.4
1 Citations from B. Keil (ed.), Aelii Aristidis Smyrnaei Quae Supersunt Omnia, Vol. 2 (Berlin,
1898, repr. 1958).
2 Aristides says he dedicated a tripod adorned with representations of Asclepius, Hygieia and
Telesphorus in the Roman-era Temple of Zeus Asclepius (Or. 50.46), which appears to have
housed offerings made primarily by elite visitors, C. Habicht, Die Inschriften des Asklepieions
(Berlin, 1969), 1314. Excavations of the Asclepieum have not, however, yielded any epigraphic
material related to Aristides, B. Puech, Orateurs et sophistes grecs dans les inscriptions dpoque
impriale (Paris, 2002) at 144; dedicatory inscriptions left by other orators of this period have
been recovered, Habicht, loc. cit. at 16 and 7180. There are parallels for the offering reported at
50.456. The dedication in Athens (IG 22 4531) of an altar to the same three divinities (a common
grouping) has been attributed to Aristides; see C.P. Jones, Three foreigners in Attica, Phoenix 32
(1978), 22234. For an example of a tripod dedicated to Asclepius commemorating artistic
performances, see IG 22, no. 3120b (Athens, C.E. 190200):
/
/

L. Robert, tudes anatoliennes (Paris, 1937), 21617 and Puech, loc. cit., 13845
consider Aristides epigram in relation to contemporary orators self-commemorative inscriptions.
3 Aside from brief remarks on epigraphic parallels or dedicatory practices, the passage receives
little comment in the annotated translations published by C.A. Behr, P. Aelius Aristides: The
Complete Works, 2 vols (Leiden, 1981), A.-J. Festugire and H.-D. Saffrey, Discours sacrs: Rve,
religion, mdecine au IIe sicle ap. J.-C. (Paris, 1986), S. Nicosia, Discorsi Sacri (Milan, 1984),
O. Schrder, Heilige Berichte (Heidelberg, 1986).
4 Aristides relatively high rate of Pindaric quotation is noted by C.A. Behr, Aelius Aristides
and the Sacred Tales (Amsterdam, 1968), 11 and n. 28, and by A. Boulanger, Aelius Aristide et la
sophistique dans la province dAsie au IIe sicle de notre re (Paris, 1923), 441. T.K. Gkourogiannis,
Pindaric Quotations in Aelius Aristides (Diss., University of London, 1999) provides a compre-

264

JANET DOWNIE

In the fourth Logos, Aristides describes the resumption of his literary training and
rhetorical performances after a year of illness.5 In spite of continuing physical
weakness, Aristides is required not just to practise a little vocal exercise, light conversation, or recitation from memory, but in fact to take up declamation in the full sense:
in the stoa near the theatre he is to perform an improvised and agonistic speech (Or.
50.15). This, he says, was the first of many such performances, for this new strength
was as if the god was providing it, and the year seemed to be one not of silence but of
training.6 In this story of professional revival, poetic composition constitutes a
distinct narrative sub-section (Or. 50.3147) in which Aristides taps the traditional
springs of literary inspiration.7 From his first tentative lines, written after falling ill
on the journey to Rome (Or. 50.31), to the public (
, Or. 50.43) choral
performances he presents in Pergamum, Aristides enjoys continual divine guidance,
and he begins and ends with Pindar as his model. In his initial paean to Apollo,
inspired in a dream by the god of poetry himself, the opening line
(Or. 50.31) echoes the beginning of Pindars second
.8 His account closes on a similarly Pindaric note:
Olympian:
the narrative sequence culminates in dream-instructions to set up a choregic
monument, partly as a mark of gratitude to the god, and partly as a memorial to the
choral performances I had given.9 The inscription for this memorial tripod crowns
the excursus on his poetic career (Or. 50.45):

hensive study of the citations in context. Pindar is by no means an exclusively epinician poet,
particularly for Aristides who, as Gkourogiannis shows, was thoroughly familiar with the entire
Pindaric corpus, much of which is lost to modern readers. Themes of crucial interest to Aristides
including Pindars self-definition as a poet and his relationship to the divine appear across the
corpus.
5 Or. 50.14. Behr (n. 4), 26, n. 19 dates the beginning of Aristides sojourn at the Pergamene
Asclepieum and hence his return to rhetorical practice to the summer of C.E. 145, approximately a year after his return from Rome in poor health (cf. Or. 48.7, 4649, 70); cf. C.A. Behr,
Studies on the biography of Aelius Aristides, ANRW 2.34.2 (1994), 11401233, at 1155, n. 58.
6

(Or. 50.18). Asclepius imposes a regimen of

literary training, alongside prescriptions for physical health, and rhetorical feats themselves may
bring about cures: e.g. Or. 50.1418, 22, 2930. On contemporary notions about how vocal
exercise could affect health, see M. Gleason, Making Men (Princeton, 1995), Ch. 4, and
A. Rousselle, Parole et inspiration: le travail de la voix dans le monde romain, History and
Philosophy of the Life Sciences 5 (1983), 12957.
7 The account of Aristides poetic training is part of the overarching narrative of his return to
professional practice (see also n. 25, below). For accounts of the place of poetry in oratorical
education of this period see Boulanger (n. 4), 427 and E. Bowie, Greek sophists and Greek
poetry in the Second Sophistic, ANRW 2.33.1 (1989), 20958.
8 O. 2.1. Cf. Bowie (n. 7), 21415.
9

(Or. 50.45) On the kind of performances that might have led up to the dedication of the tripod, see Bowie (n. 7) at 216.

A P INDARIC CHARIOTEER

265

And the following elegy had been prepared by me:

The poet, president of the contests, and chorgos himself,
has dedicated to you, lord, this memorial of choral performance.
Then there were two other verses in addition to these, one of which contained my name,
and the other the fact that all this took place with the gods guidance. But the god carried
off the victory. For on the day when the dedication was supposed to take place, on this day,
so it seemed, or a little before then, around dawn or even earlier, a divine epigram came to
me that went like this:
Not unknown among the Greeks, Aristides dedicated [this],
Illustrious charioteer of ever-flowing tales (muthoi).
I dreamed that I was inscribing this and also that I was going to dedicate it as a votive
offering precisely as if to Zeus.10

Scholars have drawn attention to the fact that Aristides name is highlighted in the
gods revised dedication.11 The inscription can then be adduced as another example of
the sense of self-importance that led certain rhetoricians of this period to identify
themselves by personal name only, without demotic or patronymic.12 However,
Aristides tells us that his name would have appeared in the first dedication as well, in
lines he does not quote here. There is no way of knowing what form this signature
would have taken, but if Aristides aim was to highlight the appearance of his name in
the gods version of the inscription, he need not have mentioned that it was included
in the original dedication as well. More striking is the shift from a simple delineation
of his responsibilities in mounting choral productions, to an impressive metaphorical
evocation of his excellence. The first epigram follows the commemorative conventions
of choregic monuments, which enumerate the names and roles of the individuals
responsible for various aspects of the choral production.13 In the second, inspired
version, the claim to official functions is replaced by a bold statement of Aristides
) reinforced by Homeric
exceptional brilliance and fame (
) and a description of his compositions as everlasting (
). All
language (
this is animated by the metaphor emphatically late in the second line of Aristides
.
as a charioteer,
Chariot racing as shorthand for excellence was current in epigraphic monuments to
literary and artistic accomplishment. From among a number of examples ranging
considerably in both date and subject matter,14 a choregic epigram of the Hadrianic
period from Athens provides perhaps the closest parallel:
10

In order to be sure that he has completely fulfilled this part of the dream-prescription (
) Aristides eventually makes a second dedication to Olympian
Zeus (Or. 50.46), probably at a temple of Olympian Zeus on his ancestral lands in Mysia. For
this temple, see Or. 49.41, 50.48, 50.1, 51.10; Robert (n. 2), 20722 locates it at the modern day
town of Alibey, north of Omerky. In the narrative context of a victory memorial, the reference
also recalls the games held in Zeus honour at Olympia, whose athletic victors Pindar celebrated.
11 Puech (n. 2), 1445 and 399400; Robert (n. 2), 21617; Habicht (n. 2) at 75.
12 A choice example is cited and interpreted by Puech (n. 2) at 399: dedicating a statue of
Demosthenes in the Pergamene Asclepieum (after dream-instructions from the god), the orator
Polemon identifies himself by personal name only, but specifies the patronymic and deme name
of the (presumably more famous) fourth-century dedicatee.
13 In this case, Aristides fulfills all the roles: poet, chorgos and agonothete (
).
14 G. Kaibel, Epigrammata Graeca (Berlin, 1878), no. 39. 3 (Athens, Dipylon, fourth century
B.C.E.):
; no. 498. 2 (Boeotian Thebes, c. third century B.C.E.):
. T. Preger, Inscriptiones Graecae Metricae (Leipzig, 1891; Chicago, 1977), no.
10 (Athens, fourth century B.C.E.):
.

266

JANET DOWNIE

(?)
[For Praxag]oras charioteered (

) an [excellent] chorus. (IG 22 3117)15

Aristides revised dedication fits well within this metaphorical and commemorative
tradition; Asclepius eloquent charioteer of ever-flowing muthoi gives poetic clat to
the responsibilities for composition and performance that Aristides had already
(Or.
enumerated in more pedestrian fashion.16 The gods victory
50.45) is partly a stylistic one that illustrates the practical aid Aristides says was
bestowed upon him so abundantly in dreams.17 But as this climax of poetic commemoration and choral performance brings us full circle from Aristides beginnings as a
Pindaric poet at Rome (Or. 50.31), we should also consider the charioteering
metaphor from a Pindaric perspective, asking what it tells us about Aristides
relationship with his divine literary editor.18
Chariot imagery is prominent in the epinician odes, particularly in places where
Pindar reflects upon his own craft.19 He uses it to evoke two broad themes: (1) an
association with gods and the divine realm; (2) craftsmanship or technical skill. In
Olympian 1, written to celebrate the victory of Hieron of Syracuse in the single horse
race of 476 B.C.E., the emphasis is on divine favour. When Pindar forecasts the
possibility of racing victories in Hierons future he also anticipates his own role in
celebrating them by alluding to his poetry in terms of a chariot metaphor:

A tutelary god keeps watch over your endeavours, Hieron, making this his concern. And as long
as he does not suddenly desert you I hope to celebrate an even sweeter victory with the swift
chariot, having found a helpful road of words (O. 1:109110).

The swift chariot is both the literal vehicle of Hieron as victor, and the metaphoric
vehicle of Pindars poetic excellence. In both cases, divine favour opens the path of
victory clear and unobstructed. Human craft also contributes to excellence, however.
15 Text from SEG 51.209; cf. S. Follet and D. Peppas-Delmouzou, The Greek East in the
Roman Context (Helsinki, 2001), 95117. The individual named here is probably the chorgos.
16 Cf. Bowie (n. 7) at 217.
17 See Or. 50.256.
18 Because the narrative context of the chariot image in the fourth Logos points towards
Pindar, I set aside another obvious precedent: the charioteer of the soul in Platos Phaedrus. In
his Or. 2 Aristides turns Platos discussion of madness and inspiration in the Phaedrus towards a
defence of oratory as a divinely inspired art, and in Or. 28 he evokes the image of a winged
chariot to describe his own performances (Or. 28.11415); cf. 28.143 for direct reference to the
Phaedrus. On the centrality of the Phaedrus in contemporary literary culture, see M.B. Trapp,
Platos Phaedrus in the Second Century, in D.A. Russell (ed.), Antonine Literature (Oxford,
1990), 14173, at 141, 1523 and 1667 for details on Aristides. Platonic chariot imagery speaks
to the theme of divine influence, but it does not engage questions of the close relationship
between inspiration and technical skill, in the way that the Pindaric chariot does. Furthermore, as
epinician poet par excellence and one of Aristides preferred literary models (see n. 4 above),
Pindars example is directly relevant to the contexts of choral performance, competition and
poetic commemoration that are at issue in the passage under consideration here.
19 D. Steiner, The Crown of Song (London, 1986). M. Simpson, The chariot and the bow as
metaphors for poetry in Pindars Odes, TAPA 100 (1969), 43849.

A P INDARIC CHARIOTEER

267

So at Olympian 9.801 Pindar hopes that his own skill makes him worthy of the
/
chariot of the Muses:
(Skilful with words may I be fit to be carried up in the chariot of the Muses).20 The
potential of the chariot metaphor as an expression of craftsmanship comes most
insistently to the fore in cases where Pindar shifts attention away from the vehicle and
towards the charioteer who drives it. In a context quite separate from poetry, a boxing
trainer responsible for moulding the innate strength and gifts of a young competitor is
presented in these terms:
< >
and I would call Melesias equal in swiftness to a dolphin through the sea, charioteer of hands
and strength. (N. 6:646)

Described as a charioteer, and likened to a dolphin in speed, the trainer Melesias

embodies the conjunction of learned skill and natural physical strength that he
nurtured in the victorious young Alcimidas.21 This potent combination of innate gift
and acquired skill returns in a remarkable image of poetic excellence from
Isthmian 7:

and mortals forget what does not reach the glorious pinnacle of wisdom, yoked to renowned
streams of verses. (I. 7.1719)

In the economy of the metaphor, attaining wisdom through poetry requires bringing
together the discipline of the yoked chariot and the spontaneous, natural inspiration
of flowing waters.
Inspiration and craft are the twin requirements of excellent performance in
Pindars world whether athletic or poetic and the image of the chariot can, as we
have seen, illuminate both. However, in the two places where he names his subordinates in the process of poetic production, Pindar uses the analogy of the
hierarchical relationship between the victor and his charioteer to figure his own
superior position as poet with respect to the chorus leader who presents his works.22
In Olympian 6, Pindar describes himself as drawing inspiration for his song from the
spring of the nymph Metope (O. 6.824) but then urges the chorus-leader Aeneas to
) his companions in performance (O. 6.878). Aeneas is thus given
spur on (
the enabling role of Phintis, Hegesias charioteer, who was in fact conscripted
(metaphorically) into poetic service earlier in the poem, when Pindar exhorted him:
20 For the chariot of the Muses, cf. I. 8.61 and Paean 7 (fr. 52h) 1314, which Aristides quotes
when he produces an example of poetic locution at Or. 45.13:
. I thank Professor
Ewen Bowie for this reference.
21 The same two metaphors (dolphin and charioteer) appear in fr. 140 B, where they seem to
describe Pindars poetry and the poetry of one of his predecessors respectively. W.J. Henderson,
Pindar Fr. 140B SnellMaehler: the chariot and the dolphin, Hermes 120 (1992), 14858 reads
the metaphors here in terms of an opposition between inspiration and craft.
22 For discussion of the parallel Pindar constructs between his own chorus leader and the
charioteer who drove for the victor see N.J. Nicholson, Aristocracy and Athletics in Archaic and
Classical Greece (Cambridge, 2005), Chs 3 and 4.

268

JANET DOWNIE

O Phintis, yoke at once the strong mules for me, as quickly as possible, so that we may drive our
chariot on a clear path and I may come to his familys very lineage (O. 6.225).

Likewise in Isthmian 2, when Pindar instructs his chorus-leader Nicasippus to impart

) to his guest-friend Thrasybulus (I. 2.478),23 he echoes the
these words (
- dispense; direct) he used earlier in the poem to describe the role of
language (
Xenocrates driver, Nicomachus, in the racing victory (I. 2.202).24 In these examples,
then, the metaphor of the charioteer creates distance if not precisely opposition
between inspiration and craft.
When Aristides borrows the metaphor of the Pindaric charioteer for the
culminating episode of his poetic itinerary in the Hieroi Logoi, he invokes a complex
web of associations: divine inspiration, skilled craft, and also a suggestion of hierarchical distance between the two. In the epigraphic tradition, as we have seen, the
image of the charioteer is a motif of honour and status. Indeed, Aristides is not
carried up passively by the Muses chariot to a divine realm; rather he is himself the
). At the same time, the motif of divine
glorious charioteer (
sponsorship is as closely entwined in his self-representation as a poet as it is in his
Pindaric model. For whose words are these that Aristides composes, publishes and
performs? The gift of the revised epigram recapitulates the crucial role of Asclepius
and other divinities in Aristides poetic development a reminder that the
) that Aristides charioteers come from the god.
ever-flowing tales (
When Aristides declares that the god carried off the victory

the statement applies not to this particular interaction only; it sums up the story of his
progress as a poet (Or. 50.3147) and, by extension, his whole professional revival.25
Without relinquishing the claims to personal literary glory that the chariot image
suggests, Aristides points at the same time to his own subordination to the god.26 This
refinement need not ultimately detract from Aristides prestige: here, as so often in the
Hieroi Logoi, we see Aristides astutely negotiating the boundary between self-aggrandizement and glorification of Asclepius. The Pindaric image of the charioteer allows
him to achieve both purposes.
In his Hymn to Asclepius (Or. 42) Aristides is explicit about his pointed interest in
Pindar as a model for the relationship between writer and god.27 Here Aristides
23
24

Son of the now-deceased Xenocrates.

25 At the end of the excursus of Or. 50.3147, Aristides makes a transition back to the subject
of oratory proper by rephrasing the compliment the god paid to him as a poet
(Or. 50.45) in more general terms:

it seemed in
every way necessary to cling to oratory (logoi) since the god had called my words (logoi)
everlasting (Or. 50.47).
26 Compare the metaphor at Or. 43.26, To Zeus, where Zeus in his role as the universal
directing power is compared to the
, while other beings (humans and other gods)
assume the enabling role of the
:

< >
And everything everywhere is full of Zeus, and for all he presides over every deed, as
teachers do with pupils and parabateis with charioteers.
27 Or. 42 postdates the Hieroi Logoi, to which it contains a reference at Or. 42.10.

A P INDARIC CHARIOTEER

269

describes his speeches as divine gifts, and he contrasts his understanding of his own
oratorical accomplishments with a story told about Pindar that the god Pan himself
performed one of Pindars paeans. The reverse is true in Aristides case, he says:
>28

<

.
In fact you [Asclepius] reversed Pindars situation. For in his case, Pan danced his paean, so the
story goes. Whereas I, if I may say so, <thought it right> to be the interpreter (
) of the
<speeches you taught>. Since you yourself [Asclepius] directed me towards them [i.e. rhetorical
studies,
] and established yourself as the commander of my training. (Or. 42.12)29

Although the syntax is disturbed,30 Aristides seems to describe himself as the

interpreter, the actor (
) of Asclepius, who directs his training and
performance in public speaking.31 In both images one from the theatre and one from
the race-track the god is the power behind Aristides accomplishments and thus the
true victor.
In conclusion, I have tried to show that the story of the revised commemorative
epigram in the fourth Hieros Logos is a more finely calibrated piece of self-promotion
than scholars have previously recognized. When the description of Aristides is
changed from poet, judge and chorgos to charioteer, what is highlighted if we read
through a Pindaric lens, is Aristides relationship with Asclepius. Figuring both divine
inspiration and skilled craftsmanship, the Pindaric language of chariot racing
illuminates the dynamic connection between god and human that defines Aristides
self-presentation as a writer of poetry, as it does his conception of his wider rhetorical
vocation.32
Princeton University

JANET DOWNIE
jdownie@princeton.edu

28 Keil (n. 1), 338 follows previous editors in positing a lacuna and offers this conjecture in his
apparatus. For other conjectures, see Keil ad loc. and G. Dindorf, Aristides, 3 vols (Leipzig, 1829)
1.68, n. 3 (Or. 6 = Or. 12 K). Behr (n. 3) at 249 and 463 translates Reiskes emendation:
<
>
I say that I am the actor of your compositions.
29 The Vita Ambrosiana of Pindar records the tradition that Pan was seen between Cithaeron
and Helicon singing (
) one of Pindars paeans, A.B. Drachman, Scholia Vetera in Pindari
Carmina, 3 vols (Leipzig, 1903), 1.2.2. For Pindars song of gratitude cf. fr. 95 and Gkourogiannis
(n. 4), 110. Aristides refers again to this incident in his oration In Defence of the Four,
F.W. Lenz and C.A. Behr (edd.), P. Aelii Aristidis Opera Quae Exstant Omnia Volumen Primum
(Leiden, 1976), at Or. 3.191, where it contributes to an argument about philotimia, a topic of
ongoing concern to Aristides; cf. Or. 28.55, where Aristides refers to O. 2.868 and invokes
Pindar as a model for poetic pride. On this oration, see I. Rutherford, The poetics of the
paraphthegma: Aelius Aristides and the decorum of self-praise, in D. Innes, H. Hine and
C. Pelling (edd.), Ethics and Rhetoric: Classical Essays for Donald Russell on his Seventy-fifth
Birthday (Oxford, 1995). On Aristides use of the Pindaric biographical tradition see Gkourogiannis (n. 4), 11416 and 11925.
30 See n. 28, above.
31 Cf. Or. 50.18, Or. 50.26.
32 I would like to express my gratitude to Professors Christopher Faraone, Shadi Bartsch and
Elizabeth Asmis for their encouragement, and for their advice on drafts of this paper. I would
also like to thank William Bubelis for enthusiastically fielding questions on epigraphic matters. I
am grateful to Professors Mark Payne and Ewen Bowie for responses to earlier presentations of
this material, and I have benefited from the comments of CQs anonymous referee.

Classical Quarterly 59.1 270293 (2009) Printed in Great Britain

270
SHORTER NOTES

SHORTER NOTES
A NEW COMIC FRAGMENT (ARISTOPHANES?)
ON THE EFFECT OF TRAGEDY
Olympiodorus, In Plat. Gorg. 33.3, p. 172, 623 Westerink, reads as follows in the
codex unicus, Marc. gr. 196 Z (M):
1

Olympiodorus account of Platos rejection of drama (Rep. 3, 394b398b, and 10,

603b606d) may partly depend on Proclus (cf. In Plat. Remp. 2.49.1319 Kroll ), and
his phrase
can be compared with
at In Plat. Remp. 1.44.27 Kroll. However, he includes other material. We
may compare Theophrastus definition of tragedy as a
2 The
and of epic as a
similar phrasing suggests that Theophrastus was Olympiodorus ultimate source. This
is supported by the fact that Theophrastus had a theory of catharsis that applied to
;3 by this term he meant both music and poetry, as did everyone down to
Philodemus De musica.4 Theophrastus theory, adapted from Aristotle, was in turn
borrowed by the Stoic Diogenes of Babylon, as we learn from Philodemus.5
The ungrammatical and odd expression
is surely an unnoticed quotation from comedy, as Westerink
suspected. Most of two comic trimeters can easily be restored, as follows:
< >
< >

<

>.

1 This is my emendation for Ms

, an error based on the anticipation of the three
occurrences of this word soon afterwards; cf. Ath. 11.505c,
2

F 708 Fortenbaugh.
Cf. F 71921 Fortenbaugh.
4 The new Bud edition by D. Delattre (Philodme de Gadara, Sur la musique, livre IV [Paris,
2007], 2 vols) finally makes fully accessible not only Philodemus views on music, but the On
music of the Stoic Diogenes of Babylon, which he summarizes at length and which cited
Theophrastus (F 7201 Fortenbaugh, now cols 812 Delattre, where the text is very different).
5 Cf. R. Janko, A first join between PHerc. 411 + 1583 (Philodemus, On music 4): Diogenes of
Babylon on natural affinity and music, CErc 22 (1992), 1239.
3

271

SHORTER NOTES

6 These verses are otherwise

The second verse may have ended with
or
unknown. Could they be part of Aristophanes pervasive commentary on the rival
genre, to which literary critics like Aristotle would owe so much? He often uses
of the audience,7 and the versification is consistent with his style.8 Alternatively, they could derive from a play that considered tragedy at length, like
Antiphanes
.9 Olympiodorus could have obtained the quotation via Theophrastus, who also wrote on comedy.10

University of Michigan

RICHARD JANKO
rjanko@umich.edu
doi:10.1017/S00098388090000214

6 Cf. Cratin. Pyt. fr. 203 KasselAustin,

; Men. Sent.
538,
< >
{
}
; Com. Adesp. fr. 1209,2 Kock = TrGF
Adesp. 26 Snell,
/
.
7 Cf. Ach. 496, Ran. 2, 132 (nom.); Pax 658 (acc.); Pax 964 (gen.); Vesp. 59, Eccl. 888, Plut. 798
(dat.); cf. Vesp. 1287, 1475, Pax 543, Nub. 518, Ran. 926, 1110, 1475, and in the singular Eq. 327,
704, Ran. 16. Antiphanes uses it in his
in the dative plural (fr. 189,16 KasselAustin),
and it is also in Adesp. com. 276,2.
8
falls in the same metrical sedes at Eq. 9, and
at Eq. 1180, Men. Asp. 394.
fills this sedes at Ach. 412, Vesp. 1511, Pax 148, Thesm. 450, Ran. 1120, fr. 392,1
KasselAustin; it is elsewhere at Ach. 400, 464, Av. 101, 1444, Ran. 90, 95, 798, 802, 834, 862, Lys.
138, Plut. 423, Men. Sicyon. 264 and Com. Adesp. fr. 1051,1 KasselAustin, and in other metres
at Eq. 401, Ran. 913, 935.
9 Cf. fr. 189 KasselAustin.
10 F 70911 Fortenbaugh.

ZENODOTUS TEXT OF HESIOD

Zenodotus of Ephesus was the first librarian in Alexandria and active as a literary
scholar in the early decades of the third century B.C.E. Best known for his muchreviled
of Homer, Zenodotus also produced an innovative alphabetical
glossary (
) and worked on the texts of poets, including Pindar, Hesiod and
Anacreon. While citations of Zenodotus readings by later Hellenistic and Roman
writers reveal much about his scholarship on the Iliad and the Odyssey (over 400
readings of his are preserved), little evidence remains of his work on Hesiods poetry.1
In fact, only once do the Hesiodic scholia provide information about the readings of
his text.2 A single thirteenth-century manuscript (Marc. gr. 464), in the hand of
Demetrius Triclinius, contains this unique comment: ad Hes. Th. 5
:

In the first modern edition of the

1 On Zenodotus Hesiodic studies, cf. G.J.C. Muetzell, De emendatione Theogoniae Hesiodeae
libri tres (Leipzig, 1833), 281; C. Gttling, Hesiodi Carmina, (Gotha, 18432), lxvilxvii; H. Flach,
Glossen und Scholien zur hesiodischen Theogonie mit Prolegomena (Leipzig, 1876), 11011;
F. Jacoby, Hesiodi Theogonia (1930), 468, 745; J. Schwartz, Pseudo-Hesiodeia: Recherches sur la
composition, la diffusion et la disparition ancienne duvres attribues Hsiode (Leiden, 1960),
2801, 614; N.A. Livadaras,
(Athens,
1963), 356; R. Pfeiffer, History of Classical Scholarship: From the Beginnings to the End of the
Hellenistic Age (Oxford, 1968), 117; K. Nickau, RE 10a (1972), 22, 38; and M.L. West, Hesiod:
Works and Days (Oxford, 1978), 64.
2 ad Hes. Th. 116 c (Di Gregorio) attributes an explanation of
to a Zenodotus, but this
is evidence for exegesis, not a text, and at any rate it is not certain Zenodotus of Ephesus is meant.

272

SHORTER NOTES

Hesiodic scholia (181423), Thomas Gaisford printed the reading of the manuscript,
.3 A decade after the publication of Gaisfords edition in
Germany, Wilhelm Muetzell suggested emending
to
.4 Ten years later Karl
Gttling followed Muetzells suggestion, with the specification that
be
understood with
.5 The emendation was taken up by Hans Flach in his edition of
the scholia to the Theogony where he maintained that kdsesi should be understood
rather than Gttlings
.6 Lamberto Di Gregorio followed Flach in the
most recent edition of the Theogony scholia.7 Muetzells emendation of
for
is predicated on understanding that Zenodotus text of Hesiods poems was called by
later scholars either a
or an
, though there is no evidence for this. A
third possibility, one that would support the scholions reading in the manuscript, is
that ancient scholars referred to Zenodotus copy of Hesiods poems as an
. In fact, the Homeric scholia twice use the phrase
/
to refer to Zenodotus text of Homers Iliad as though
is meant: A ad Hom. Il. 13.808 a (Erbse)
<
>:
; and A ad Hom. Il. 19.26 a (Erbse)

.8 These parallels
confirm that
(sc.
) is the correct reading for the
scholion in Marc. gr. 464 and discredit Muetzells emendation.
What this scholion preserves for us is the precious information that Zenodotus read
( )
for
in his text of Hesiods Theogony. A number of late
medieval manuscripts of the Theogony (for example, Vat. gr. 915, Laur. conv. suppr.
158, Paris gr. 2833) still retain Zenodotus variant
( )
or a form of it. There
are at least three explanations for the origin of Zenodotus reading
( )
.
One possibility is that the confusion between
( )
and
reflects
some copyists spelling error since
and
are orthographically quite similar.
Another solution was offered by Felix Jacoby, who imagined that Zenodotus had
discovered the form
( )
in an ancient text of the poem (in libris
antiquioribus) and included it in his own copy.9 But Martin West has provided the
most convincing explanation for the different names when he observed that the
scholion records alternative spellings of the rivers name as reflected in two different
Greek dialects: If the initial consonant represents an original labio-velar, - will be
correct for Boeotia, while Attic and koine would have -.10 In other words, Hesiod
and speakers of Boeotian and other Aeolic dialects would call the stream on Mt
Helicon Permessos, while speakers of Attic and related dialects (such as Ionic)
would refer to the same stream as the Termessos.11
3

T. Gaisford, Poetae Minores Graeci, 2 vols (Oxford, 1814; Leipzig, 1823), 2.463.
Muetzell (n. 1), 281.
Gttling (n. 1), lxvi.
6 Flach (n. 1), 111, 209.
7 L. Di Gregorio, Scholia vetera in Hesiodi Theogoniam (Milan, 1975), 4.
8 Eustathius (on Iliad 2.568, 289.38) offers a similar phrase:
; cf.
Pfeiffer (n. 1), 117, n. 5; K. Nickau, Untersuchungen zur textkritischen Methode des Zenodotos von
Ephesos (Berlin and New York, 1977), 5, n. 16; and M.L. West, Studies in the Text and Transmission of the Iliad (Munich and Leipzig, 2001), 55, n. 23.
9 Jacoby (n. 1), 75.
10 M.L. West, Hesiod: Theogony (Oxford, 1966), 153.
11 On the changes of initial labio-velars into
- or - before front vowels, cf. C.D. Buck, The
Greek Dialects (Chicago, 1955), 612; R. Schmitt, Einfhrung in die griechischen Dialekte
(Darmstadt, 1977), 6970, 76, 81.
4
5

SHORTER NOTES

273

A recent theory may confirm Wests dialectal explanation for the spelling
( )
in the text of Zenodotus, and allow for a clearer understanding of his
copy of Hesiods Theogony. In his study of the transmission of the Iliad, West has
suggested that Zenodotus brought to Alexandria from his native Ephesus an Ionian
rhapsodes copy of Homers poems which he used as the base copy for his
.12 According to Wests model, Zenodotus notated the rhapsodes copy of
the Iliad and Odyssey with critical marks but without altering the underlying text,
perhaps including variant readings in its margins. An important clue to the origins of
Zenodotus text is the inclusion of various neo-Ionic or hyper-Ionic forms, which he
took as evidence of its production in Ionian surroundings.13 Wests suggestion has met
with enthusiastic, if not universal, acceptance.14 Wests proposed solution is simple
and elegant, and neatly explains the puzzling and seemingly arbitrary readings in
Zenodotus text of Homer.
If while still a young man in Ephesus Zenodotus had procured and annotated a
copy of Homer which he later brought with him to Alexandria, we can well imagine
that he did the same with a personal copy of Hesiods poems. Zenodotus text of
Hesiod could thus have been a rhapsodic exemplar which contained the AtticIonic
spelling
( )
, reflecting what was actually recited in Ephesus and thus
differing from the AeolicBoeotian
. If this is the case, a rhapsodic
exemplar of Hesiod which shows dialectal traces of its production in Asia Minor
would have been the base copy for Zenodotus text of the Theogony. While this is
only a hypothesis, it explains the variant dialectal spelling attributed to Zenodotus by
the scholia to the Theogony, and accords well with the most satisfactory theory about
his
of Homer.
If the suggestion that a rhapsodes copy of the Theogony found its way from
Ephesus to Alexandria is accepted, further evidence of the Ionian background of
Zenodotus text comes to light. The medieval manuscripts of the Theogony transmit
the Attic form of the pronoun
in line 126, but a second-century citation
(Theophilus Apol. Ad Autolycum 2.6.11) and a fourth- or fifth-century papyrus (P.
Achmm 3) contain instead the neo- or hyper-Ionic form
.15 A papyrus fragment
of the Catalogue of Women (fr. 45 MW = fr. 37 Hirschberger) contains the similar
form
. The Hesiodic scholia are silent on whether
or
stood in the
Alexandrian copies of Hesiods Theogony (and there are unfortunately no scholia to
the Catalogue of Women), but the scholia to Homer indicate that Zenodotus read the
neo- or hyper-Ionic form
in his
of the Iliad: A ad Hom. Il. 14.162
b (Erbse)
:

.16 The neo- or hyper-Ionic form in Zenodotus

12 West (n. 8), 3345; id., Zenodotus Text, in F. Montanari (ed.), Omero tremila anni dopo
(Rome, 2002), 13742.
13 West (n. 8), 43.
14 F. Montanari, Alexandrian Homeric philology. The form of the ekdosis and the variae
lectiones, in M. Reichel and A. Rengakos, Epea Pteroenta: Beitrge zur Homerforschung
Festschrift fr Wolfgang Kullmann zum 75 Geburtstag (Stuttgart, 2002), 11940, at 123; R. Janko,
Seduta di Chiusura, in Montanari (n. 12), 65366, at 658; id., review of West (n. 8), in CW 97
(2003), 1001, at 100. A. Rengakos argues against the hypothesis in his review of West (n. 8), in
BMCR 2002.11.15: http://ccat.sas.upenn.edu/bmcr/2002/2002-11-15.html.
15 West (n. 10), 81 argues that the correct reading is
. Cf. Buck (n. 11), 7980; Schmitt (n.
11), 103.
16 Cf. Aim ad Hom. Il. 1.271 a (Erbse): <
:>

274

SHORTER NOTES

of the Iliad is the same as the variant dialectal form in Theogony 126 and
the reading of the papyrus of the Catalogue of Women. We might therefore suspect,
and with good reason, that the neo- and hyper-Ionicisms that existed in the texts of
Homer and Hesiod stem from copies produced by rhapsodes in an Ionian setting, and
that
( )
,
and
all once appeared in Zenodotus copy of
Hesiods poems. The methods by which Zenodotus readings have passed into papyri
of the Theogony and the Catalogue, medieval manuscripts of the Theogony and the
Hesiodic scholia are opaque; yet taken together they point to the fact that ancient
scholars who came after Zenodotus took an interest in preserving his readings and
thoughts if only to disagree with them. This theory about an Ionian rhapsodic copy
of Hesiod in third-century Alexandria has the additional conclusion that Zenodotus
text of Hesiod contained both the Theogony and the Catalogue of Women.17 This may
serve as another reminder that ancient views on the authenticity of Hesiods poems
differ from our own.
Cornell University

C.M. SCHROEDER
chad.matthew.schroeder@gmail.com
doi:10.1017/S00098388090000226

17 Schwartz (n. 1), 2801 hints in this direction, following the suggestion about the Ionian
background of Zenodotus text of Homer made by G. Pasquali in Storia della tradizione e critica
del testo (Florence, 1934), 2401.

THE DISUNION OF CATULLUS FRATRES

UNANIMI AT VIRGIL, AENEID 7.3356
That Virgil presents the Latin War of Aeneid 712 as, among other things, a civil war
between proto-Romans, has been noted by many readers of the poem.1 The language
of civil war becomes prominent in his account of the outbreak of war in Book 7. Here
it is the typology of familial discord as a reflection of civil discord that Virgil employs
most conspicuously. At 7.32340 Juno commissions the Fury Allecto, the embodiment of familial strife (odit et ipse pater Pluton, odere sorores | Tartareae monstrum,
7.3278), to stir up discord between eventual son-in-law Aeneas and father-in-law
Latinus. Junos pitting of these two against one another (hac gener atque socer coeant
mercede suorum: | sanguine Troiano et Rutulo dotabere, uirgo, 7.31718) recalls and
so thematically prefigures a later Roman civil war, that between father-in-law Caesar
and son-in-law Pompey, a conflict highlighted by Anchises in the previous book
(aggeribus socer Alpinis atque arce Monoeci | descendens, gener aduersis instructus Eois,
6.8301).
Warring fathers and sons, we learn from Juno, are to be accompanied by battling
brothers. At 7.33540 Juno gives Allecto her formal assignment:

1
On civil war in the Aeneid, see e.g. S.J. Harrison, Virgil as a poet of war, PVS 19 (1988),
4868, esp. 636; F. Cairns, Virgils Augustan Epic (Cambridge, 1989), 85108; P. Hardie, Tales
of unity and division in imperial Latin epic, 5771, in J.H. Molyneux (ed.), Literary Responses to
Civil Discord (Nottingham, 1993); and N.M. Horsfall, A Companion to the Study of Virgil
(Leiden, 1995), 15561, with further bibliography.

SHORTER NOTES

275

tu potes unanimos armare in proelia fratres

atque odiis uersare domos, tu uerbera tectis
funereasque inferre faces, tibi nomina mille,
mille nocendi artes. fecundum concute pectus,
dissice compositam pacem, sere crimina belli;
arma uelit poscatque simul rapiatque iuuentus.
You have the power to arm brothers of one spirit for battle, and to overturn homes with hatred.
You are able to strike whips and funeral torches upon houses. You have a thousand names, and a
thousand talents for doing harm. Shake your fertile breast, dislodge the agreed-on peace, sow
the grounds for war. May the youth wish for arms, at the same time demand them, and snatch
them up!

So Allectos task is to send the Latins and Trojans brothers in that they all have
Italian ancestors,2 and that they all are Roman forefathers headlong into war with
each other. With her attacks on Amata (7.341405), Turnus (7.40674), and then the
hounds of Ascanius (7.475504), the Fury of course succeeds.3
The effectiveness of Virgils presentation of Allecto as an agent of familial discord,
and, more broadly, of his Latin War as a fratricidal struggle, is heightened by an allusion in these lines to Catullus 9. This short hendecasyllable is addressed to Catullus
friend Veranius, who has just come home from service in Spain:
Verani, omnibus e meis amicis
antistans mihi milibus trecentis,
uenistine domum ad tuos penates
fratresque unanimos anumque matrem?
uenisti. o mihi nuntii beati!
uisam te incolumem audiamque Hiberum
narrantem loca, facta, nationes,
ut mos est tuus, applicansque collum
iucundum os oculosque suauiabor.
O quantum est hominum beatiorum,
quid me laetius est beatiusue?

10

Veranius, you who of all my friends are worth more than three hundred thousand, have you
come home, to your hearth, and to your brothers of one spirit, and to your old mother? You
have. O wonderful news to me! I shall visit you unharmed, and hear you telling about the places,
deeds, and peoples of the Hiberi, as is your custom. And, clinging to your neck, I shall kiss your
dear mouth and eyes. O, of all the more blessed men that there are, who is happier or more
blessed than I?

The poem is a celebration of homecoming and reunion. And home here is less the
physical space of Veranius house than it is the family with whom he is reunited: the
fratres and mater in line 4 are a direct extension of the domus and penates to which
Veranius returned (uenisti) in the preceding line.4
2 Time and again in the poem we read of Dardanus, and thus his descendent Aeneas, Italian
origins. Troys Italian roots are addressed most explicitly at 3.1638, 7.20511, and 7.2402. On
Virgils novel adaptation and treatment of the Dardanus myth, see V. Buchheit, Vergil ber die
Sendung Roms (Heidelberg, 1963), 16372. Another more strictly civil dimension of the Latin
War is the struggle between the exiled king Mezentius and his former Etruscan subjects.
3 Allectos stirring up of familial discord here in Aen. 7 finds parallels in the opening of
Senecas Thyestes, where the Fury goads the shade of Tantalus to perpetuate the strife between
Atreus and Thyestes. On the influence of Aen. 7.323571 on the prologue of the Thyestes, see R.J.
Tarrant, Senecas Thyestes (Atlanta, 1985), 856, n. 2.
4 So R.M. Nielsen, Catullus 9 and 31: the simple pleasure, Ramus 8 (1979), 16573, who
writes at 169: In Carmen 9 home is conceived of as much more than physical shelter for

276

SHORTER NOTES

I would like to focus on the collocation fratres unanimos in line 4, a phrase that, it
has been argued, stands at the core of Catullus poem.5 After Plautus, who uses the
older unianimus at Stichus 731,6 Catullus is the first surviving author to employ the
adjective unanimus. We see it twice more in his poetry, at 30.1 and 66.80.7 The next
attested authors to use the adjective are Virgil, who employs it a total of three times,8
and his contemporary the epigrammatist Domitius Marsus.9 Like Catullus, Virgil
modifies frater with unanimus just once, at Aeneid 7.335, where, as we saw above, the
phrase also appears in the accusative plural.
The correspondence is noted by many commentators on the two passages, but
seldom do they argue for a direct allusion.10 And, while the appearance of the phrase
fratres unanimi is unique in each of the authors, it is true that we find unanimus (and
its alternate unanimis) as an epithet for siblings, spouses and close companions
elsewhere in Catullus and Virgils poetry, and in later literature.11 But, as often, it is
the contexts in which we encounter the two phrases more specifically, the pointed
discrepancy between the contexts that establish the case for allusion. As we have
seen, Catullus 9 is a poem about Veranius reunion with his brothers, mother and his
friend Catullus. The Virgilian passage, contrarily, is about the disunion that Juno bids
Allecto bring to families and homes. Let us look again at Aeneid 7.3356:
tu potes unanimos armare in proelia fratres
atque odiis uersare domos

and at Catullus 9, 34:

uenistine domum ad tuos penates
fratresque unanimos anumque matrem?

The task that Allecto is instructed to perform is emphasized by the placement of the
words in line 335: the martial words armare and proelia literally stand between the
unanimos fratres, a figurative prolepsis of the rending apart that the Fury is about to
human life; it is a spiritual reality, one found in the bond among people enjoying the closest ties
of blood and kindred.
5 Nielsen (n. 4), at 169: The importance of human contact, of loyalty, and of verbal communication as the primary elements in Catullus presentation of homecoming in Carmen 9 is
reinforced by his central placement of the adjective unanimos. This epithet summarizes the mood
of all gathered to welcome Veranius.
6 St. 731: ego tu sum, tu es ego, unianimi sumus.
7 At 30.1, of the friends to whom the addressee Alfenus has been false (unanimis sodalibus);
and at 66.80, of the husbands of the brides addressed by the Lock of Berenice (unanimis
coniugibus).
8 At 4.8, of Didos sister Anna (unanimam adloquitur male sana sororem); and at 12.264, where
Tolumnius addresses his fellow Rutulians (uos unanimi densete cateruas | et regem uobis pugna
defendite raptum), as well as at 7.335.
9 Fr. 1 (Courtney), 12: omnia cum Bauio communia frater habebat, | unanimi fratres sicut
habere solent.
10 An exception is F.P. Simpson, Select Poems of Catullus (London, 1942 [1879]), xxxviii, who
included this correspondence in his catalogue of imitations of Catullus by Virgil. N.M. Horsfall,
Virgil, Aeneid 7. A Commentary (Leiden, 2000), 233, writes ad loc. that unanimus/-is is from
Catullus. On Virgils engagement with Catullus generally, see C. Nappa, Catullus and Vergil,
37798, in M.B. Skinner (ed.), A Companion to Catullus (Malden, MA, 2007).
11 See nn. 7 and 8 above, as well as A.S. Pease, Publi Vergili Maronis, Aeneidos Liber Quartus
(Cambridge, MA, 1935), 912, ad Aen. 4.8 for cases of unanimus/-is modifying other categories
of relatives and friends. Along with Domitius Marsus (see n. 9), Silius (13.651), and Statius
(Theb. 8.669 and 10.727) also modify frater with unanimus, perhaps under the influence of their
epic predecessor Virgil.

SHORTER NOTES

277

execute.12 The disrupted homes (domos) in line 336 are an extension of the brothers of
335. In this regard these lines share with lines 34 of Catullus 9 the assimilation of
home and family. Indeed, if we consider this close sympathetic link between lines 335
and 336, the ordering of the words too can be read as a chiastic allusion to Catullus
similarly connected lines (Catullus A domum B fratres C unanimos Virgils C
unanimos B fratres A domos). We have here, then, a type of oppositio in imitando:13
Virgil incorporates and then immediately pulls apart Catullus fratres unanimos and
the home that they represent.
Another contrast evoked by the allusion at 7.3356 is that between the homecomings lying at the heart of Catullus 9 and Aeneid 712. Veranius safe (incolumem,
line 6) return to his family and friend Catullus is one of affection (lines 89) and
unspeakable joy (lines 1011). Aeneas homecoming will be much different. Italy is his
ancestral home, where he now belongs, as Aeneas articulates in his address to the
Trojan penates at 7.1212 (o fidi Troiae saluete penates: | hic domus, haec patria est),
and as the river god Tiberinus reassures him at 8.39 (hic tibi certa domus, certi [ne
absiste] penates).14 Here he and Lavinia are to wed and establish, or re-establish, his
races domus in Italy. But so contrary to Veranius welcoming party the fratres
unanimi that await Aeneas return will be his opponents in war.
Strengthening the argument for allusion to Catullus at Aeneid 7.3356 is the
presence of other Catullan correspondences in this episode. As we saw above, Virgils
dubbing of Aeneas and Latinus as gener atque socer just above our passage at 7.317
recalls his use of the same terms for Caesar and Pompey at 6.8301. But it is at
Catullus 29.24 (socer generque, perdidistis omnia?) that we first see this slogan for
Caesar and Pompey that Virgil adopts15 and then extends into his own civil war.
Further, in the surrounding lines in Book 7 he twice incorporates language from
Catullus 64. At 7.3023, Junos complaint about the failure of various sea obstacles to
halt Aeneas (quid Syrtes aut Scylla mihi, quid uasta Charybdis | profuit) is adapted
from Ariadnes indictment of Theseus parentage at 64.156 (quae Syrtis, quae Scylla
rapax, quae uasta Charybdis). Then at 7.356 Virgil describes Allectos gradual seizure
of Amata (necdum animus toto percepit pectore flammam) with language and rhythm
drawn from Catullus account of Ariadnes falling for Theseus (64.92: cuncto concepit
corpore flammam).16
12 Horsfall (n. 10), 233, writes of this phrase that [b]rothers ought to be unanimi [H]ere
therefore the adj. almost concessive; he compares the application of concordes to Caesar and
Pompey at 6.827. Cairns (n. 1), at 101, aptly calls unanimos fratres at 7.335 a combined allusion
to fraternal concord and its opposite, civil war, which illustrates well the standard ancient equivalence of familial and public harmony. C. Bannon, The Brothers of Romulus: Fraternal Pietas in
Roman Law, Literature, and Society (Princeton, 1997), 148, similarly observes Allectos inversion
of fraternal pietas in 7.3358, but also downplays the extent to which Virgil presents his Latin
War as a fratricidal, civil war.
13 As G. Giangrande ( Arte Allusiva and Alexandrian epic poetry, CQ 17 [1967], 8597)
labels such an inversion of ones model. R.F. Thomas (Virgils Georgics and the art of reference,
HSCP 90 [1986], 17198) terms such a reference a correction, a process defined at 185: The poet
provides unmistakable indications of his source, then proceeds to offer detail which contradicts
or alters that source.
14 See n. 2 above. And on Virgils development of Italy as Aeneas patria (a theme first made
explicit at 1.380: Italiam quaero patriam), see Cairns (n. 1), 10928.
15 So M.C.J. Putnam, The lyric genius of the Aeneid, Arion 3 2/3 (19956), 81101, at 89:
Virgil draws the collocation of socer and gener [at 6.8301], and a touch of that poems irony,
from Catullus 29 (24) where the two relatives are apostrophized as incorporations of Roman
immorality. W.A. Camps, An Introduction to Virgils Aeneid (Oxford, 1969), 97, regards the
possible evocation of Caesar and Pompey at 7.31718 rather as a sub-conscious association.
16 R.O.A.M. Lyne, Further Voices in Vergils Aeneid (Oxford, 1987), 1617, explores the erotic
implications of this allusion.

278

SHORTER NOTES

Furthermore, Virgils use of Catullus 9 elsewhere in the Aeneid has been observed.
At 6.687, we find as the first word of a speech rich with Catullan allusions the verb
uenisti. As Michael Putnam has noted, Virgil uses this form of uenio only here; and
Catullus uses this form only twice, at lines 3 and 5 of poem 9, of Veranius return.17
Just as Catullus hails the returning Veranius with uenisti uenisti, Anchises at Aeneid
6.687 exclaims uenisti tandem to his son when they meet again in the Underworld.
Their reunion, however, will be brief, to be followed by a more lasting separation. The
fleeting and unfulfilling nature of Aeneas reunion with his father, Putnam argues, is
brought out by the contrast with Veranius living, lasting reunion with his family and
friends in Catullus 9.18
A similar contrast with Catullan precedent comes at 7.3356, where, as we have
seen, Virgil adopts and simultaneously dissolves Catullus unanimos fratres. The loss
of the unanim-ity that Catullus captured in his poem will define Virgils Latin War.
Junos bidding of Allecto at 7.33140 begins the dissolution; and, indeed, over the
ensuing lines of Book 7 Virgil emphasizes Allectos stirring up of hostile animi. As we
saw above, at 3546 we read of the gradual movement of Allectos poison from
Amatas body on to her animus (ac dum prima lues udo sublapsa ueneno | pertemptat
sensus atque ossibus implicat ignem | necdum animus toto percepit pectore flammam).
The Furys attack on Turnus is noted as successful when at 475 he is spreading
combative animi to his Rutulian countrymen: dum Turnus Rutulos animis audacibus
implet. Just after, at 4812, we read that Allectos infection of Ascanius hunting dogs
(who are thus able to track down Silvias stag) caused the enflaming of Latin rustics
animi: quae prima laborum | causa fuit belloque animos accendit agrestis. Then, after
war has broken out, Allecto vows at 550 to enflame more animi with the love of war,
should Juno bid her do so (accendam animos insani Martis amore). Indeed, the stirring
up and fractiousness of animi that we witness over the course of Book 7 is announced
in the books proem, where Virgil declares that he will sing of kings driven by animi to
their deaths (dicam acies actosque animis in funera reges, 7.42).19
Fratres unanimi, it is clear, have no place in this conflict. The Latin Wars impetus is
the precise lexical opposite of unanim-ity, discordia, the splitting or separation of
kindred hearts. Allecto boasts of discordias arrival in Latium at 7.545, just before her
pledge at 7.550 to enflame more animi as needed: en perfecta tibi bello discordia tristi.
And Virgils illustration of civil discord in Aeneid 7 is made more potent, more painful
by his allusion at 7.3356, where the reunion with fratres unanimi that Catullus had
commemorated is torn apart, replaced by disunion and discordia.20
College of the Holy Cross

TIMOTHY JOSEPH
tjoseph@holycross.edu
doi:10.1017/S00098388090000238

17 Putnam (n. 15), at 95. At 96 Putnam looks at further verbal connections between Aen.
6.684702 and Cat. 9.
18 Putnam (n. 15), at 96: There can be no physical contact in the world of ghosts, no possibility for living son to embrace dead father It is this notion [of the horror of separation] that
Virgil chooses to emphasize in contrast with Catullus 9.
19 Virgil keeps the opposition of proto-Roman animi fresh in our minds as we enter Book 8.
Before the relative calm of Aeneas trip to Pallanteum, we are reminded in the books opening
lines of the agitated animi of the Latins (8.4: extemplo turbati animi) and of Aeneas himself
(8.201: animum nunc huc celerem nunc diuidit illuc | in partisque rapit uarias perque omnia uersat).
20 I am thankful to Christopher Krebs, Michael Putnam, Richard Thomas, the CQ editor
Rhiannon Ash, and the anonymous reader at CQ for their helpful comments on earlier drafts of
this paper.

279

SHORTER NOTES

HORACE, ODES 2.14.14

Frustra cruento Marte carebimus
fractisque rauci fluctibus Hadriae,
frustra per autumnos nocentem
corporibus metuemus Austrum

14

In vain do we avoid the dangers of war, sea-faring, and climate, for inevitably (1721)
we must behold the Underworld and its infamous sinners.
As a rule, waves are broken in Latin quite explicitly by or on rocks, reefs, sandbanks or the like,1 and the fact that none of those is mentioned here to amplify fractis
(14) may perhaps raise doubts about our text. Add the consideration that we should
have expected a reference to the open seas stormy waves, such as encountered by a
typical traveller crossing the Adriatic, rather than an allusion to the perils of off-shore
reefs or shoals, and doubts increase.
Horace speaks elsewhere of the Adriatics notorious storms, whipped up especially
by the south winds, Notus, quo non arbiter Hadriae / maior (Odes 1.3.1416), and
Auster, dux inquieti turbidus Hadriae (Odes 3.3.45). He claims to know personally
quid sit ater/ Hadriae sinus and prays, conventionally, that only enemies wives and
children might experience caecos / motus orientis Austri et / aequoris nigri fremitum
(Odes 3.27.1819; 213). Of particular interest in these passages is his application of
the adjectives ater and niger to stormy waters, for which we may compare his mention
of a sea too stormy for fishing, atrum / defendens piscis hiemat mare (Sat. 2.2.1617),
and Virgils description of waves darkened by the north wind, fluctus atros
Aquilone (Aen. 5.2); the winds themselves sometimes are called dark or black: Odes
1.5.67, et aspera / nigris aequora ventis; Epod. 10.5, niger Eurus inverso mari; Cat.
68.63, in nigro iactatis turbine nautis; Virg. G. 3.278, nigerrimus Auster.
The association of darkness with wind-driven, stormy waters will encourage a
suggestion that, in the present line, Horace may have written not fractis but atris:
atrisque rauci fluctibus Hadriae. It may be felt that this atris is uncomfortably close to
ater in 17, uisendus ater flumine languido / Cocytus errans, but the repetition will have
been rhetorically deliberate: in vain shall we eschew the Adriatics lively dark waves for
we must soon view Cocytus funereally dark stream. While atris will entail a slight
decrease in the verses alliteration of f and c, the adjectives connotation of stormy,
noisy wind may complement the aural imagery of rauci Hadriae. As for the posited
corruption, we might suppose that, with frustra just above (13) and also just below
(15), atris will have been copied as fratris which then was corrected to the manuscripts fractis.
Brooklyn College / Penn State University

ARCHIBALD ALLEN
archibaldallen@msn.com
doi:10.1017/S0009838809000024X

1 Compare, for example, Ov. Fast. 4.282, quaeque Carysteis frangitur unda vadis; Lucan 9.308,
aequora fracta vadis; Sil. Ital. 5.398, fractasque in rupibus undas. And see TLL s.v frango, II A 2, de
fluctibus sim.

280

SHORTER NOTES

SEXUAL PUNS IN OVIDS ARS

AND REMEDIA 1
I. ARS AMATORIA 2.1212
sis licet antiquo Nireus adamatus Homero
Naiadumue tener crimine raptus Hylas,
ut dominam teneas nec te mirere relictum,
ingenii dotes corporis adde bonis.
forma bonum fragile est, quantumque accedit ad annos,
fit minor et spatio carpitur ipsa suo.
nec uiolae semper nec hiantia lilia florent,
et riget amissa spina relicta rosa:
et tibi iam uenient cani, formose, capilli,
iam uenient rugae, quae tibi corpus arent.
iam molire animum, qui duret, et astrue formae:
solus ad extremos permanet ille rogos.
nec leuis ingenuas pectus coluisse per artes
cura sit et linguas edidicisse duas.

110

115

120

Ovids instructions to his student at Ars amatoria 2.10922, to remember the

transience of his own good looks and spend time cultivating his mind in compensation, have usually been interpreted by scholars as a manifestation of the praeceptors
more general concern with cultus the civilization or refinement central to the
sophisticated urban society in which the poet sets his amatory intrigues.2 But however
seriously we take Ovids dedication to a wider philosophy of cultus, the ultimate goal
of his teachings is rarely allowed to remain latent for long: these prescriptions have a
purpose, and their aim is ultimately that which informs the entire curriculum of the
Ars amatoria the acquisition and continued enjoyment of sexual pleasure. This
passage is in fact a restatement of the famous lines in Book 1 in which the praeceptor
with mock pomposity urges recourse to the traditional syllabus of Roman rhetorical
education, but for distinctly un-Roman purposes (Ars 1.45962):3
disce bonas artes, moneo, Romana iuuentus,
non tantum trepidos ut tueare reos:
quam populus iudexque grauis lectusque senatus,
tam dabit eloquio uicta puella manus.

460

The introduction of the extended exemplum of Ulysses and Calypso at 2.123 seems to
have precluded a provocative concluding aphorism along the lines of 1.4612, and the
apparent omission here of one of Ovids characteristically facetious reminders of the
fundamental goal towards which these teachings are designed to lead may well
contribute to the impression that his advocacy of a liberal education is intended to
1 My thanks to Maria Wyke, Matthew Fox and Rhiannon Ash, to audiences in St Andrews,
Glasgow and Oxford, and to CQs anonymous referee, for helpful comments on previous drafts
of these notes.
2 On Ovid and cultus, see e.g. J.B. Solodow, Ovids Ars Amatoria: the lover as cultural ideal,
WS 90 (1977), 10627; M. Myerowitz, Ovids Games of Love (Detroit, 1985); P.R. Hardie, Ovids
Poetics of Illusion (Cambridge, 2002), 1.
3 On these lines see K. Volk, Ars Amatoria Romana: Ovid on love as a cultural construct, in
R.K. Gibson, S.J. Green and A.R. Sharrock (edd.), The Art of Love. Bimillennial Essays on Ovids
Ars Amatoria and Remedia Amoris (Oxford, 2006), 23551 at 2401.

SHORTER NOTES

281

carry a wider significance.4 But in fact the way in which Ovid characterizes the skills
his student is to acquire from this education brings us immediately back to the real
aim of this supposedly high-minded advice: nec leuis ingenuas pectus coluisse per
artes | cura sit et linguas edidicisse duas (1212). Commentators rightly point out that
the injunction to become conversant in duae linguae recalls the high value placed on
bilingualism by the educated Roman elite: the two linguae in question are of course
Latin and Greek.5 As Cicero enjoins his son Marcus at De officiis 1.1: ut ipse ad meam
utilitatem semper cum Graecis Latina coniunxi neque id in philosophia solum, sed etiam
in dicendi exercitatione feci, idem tibi censeo faciendum, ut par sis in utriusque orationis
facultate.6
Had Ovid been Cicero, he too could no doubt have used an expression similar to
utraque oratio, or some more metrically tractable equivalent.7 But Ovid is Ovid, and
he does not; the expression linguas edidicisse duas (122) here is far from fortuitous,
and its context within the Ars amatoria invests it with a crucial ambiguity. After all,
there is more than one way to learn thoroughly two tongues, and two tongues come
in handy for rather more than just public speaking (or even private reading) we
might think of the pugnantes linguae of Tibullus 1.8.37, for instance, or the probing
linguae of Amores 2.5.578 (tota labellis | lingua tua est nostris, nostra recepta tuis).
What we have here, then, is a characteristic Ovidian double entendre: by describing the
activity he is recommending in terms of the outcome it is intended to procure, the
praeceptor amoris conflates the means with the end, and in so doing keeps his
students eye firmly fixed on the ultimate goal of his amatory training.
Alison Sharrock seems to have noticed this play when she detects in this passage a
joke about seduction for the reader, and identifies a further pun in the use of
double-tongued to mean deceitful.8 But in fact linguas edidicisse duas is only the
clinching punch line of a sequence of smutty double meanings that has already begun
in the hexameter of this couplet. Ovids immediately preceding precept that the
would-be lover should make it his business pectus coluisse (121) admits a comparably
anatomical meaning. On a literal level, of course, he is urging the student of love to
cultivate the mind, pectus being the seat of thought;9 but when Propertius pictures
4 For a slightly more pragmatic reading of the lines, see however P. Murgatroyd, Ovid with
Love: Selections from Ars Amatoria I and II (Chicago, 1982), 154, on 2.1212: The cultured and
sophisticated Ovid advises his pupil to acquire similar culture and sophistication. This is yet
another instance of good advice in this section, since such an education would make the reader
more impressive in most girls eyes (and in those of their acquaintances) and should mean that he
would be more interesting, have more depth and seldom be short of conversation (in particular a
mastery of Greek would extend his reading and knowledge).
5 M. Janka, Ovid Ars Amatoria Buch 2: Kommentar (Heidelberg, 1997), 125, on 1212; A.R.
Sharrock, Seduction and Repetition in Ovids Ars Amatoria 2 (Oxford, 1994), 4950 with 50, n.
45. On bilingualism at Rome, see generally J.N. Adams, Bilingualism and the Latin Language
(Cambridge, 2003); J.N. Adams, M. Janse and S. Swain (edd.), Bilingualism in Ancient Society.
Language Contact and the Written Text (Oxford, 2002).
6 On Ovids use of De officiis elsewhere in the Ars, see M. Labate, Larte di farsi amare. Modelli
culturali e progetto didascalico nellelegia ovidiana (Pisa, 1984), 12174; R.K. Gibson, Ovid: Ars
Amatoria Book 3 (Cambridge, 2003), 223 and index s.v. De officiis and the Ars Amatoria; id.,
Excess and Restraint. Propertius, Horace, and Ovids Ars Amatoria (BICS suppl. 89; London,
2007), esp. 802, 11722, 1269, 1323.
7 For parallels in Latin verse, see Horace, Carm. 3.8.5, docte sermones utriusque linguae
(impossible in dactylic metre); Sat. 1.10.23, sermo lingua concinnus utraque; Martial 10.76.6,
lingua doctus utraque.
8 Sharrock (n. 5), 50.
9 See Janka (n. 5), 125.

282

SHORTER NOTES

Cynthia following his funeral cortge nudum pectus lacerata (Prop. 2.13.27) he is
scarcely envisaging his girlfriend inflicting mental harm on herself, and when Ovid
exclaims upon seeing Corinnas naked body quam castigato planus sub pectore uenter!
(Amores 1.5.21), it is not perhaps the quality of her education that most excites him.
So we can see that at Ars amatoria 2.1212 the elegiac magister insinuates that his
pupil should take it upon himself to cultivate10 not just (his) mind, but (someone
elses) breast.11 In conjunction with the instruction to acquire an intimate familiarity
with two tongues, it becomes clear that there is more to this concluding couplet than
a disinterested call to self-improvement and if, with Sharrock, we see in ingenuas
artes (121) a nod in the direction of Ovids own Ars,12 we should expect nothing less.
The praeceptors sly, suggestive wit is as active as ever, and yet another series of
instructions is brought to a neat climax (as it were) with outrageous epigrammatic
humour.
II. REMEDIA AMORIS 75766
Much of Ovids Remedia amoris is concerned with picking up and countering
suggestions made in the Ars amatoria for the successful pursuit of love.13 One
prominent instance of this occurs towards the end of the book, where the speaker
revisits the reading-list he had given to his female disciples at Ars 3.32946.
Unsurprisingly, the register of proscribed literature here includes much of the
material we find recommended in the previous passage, described in very similar
terms:14
eloquar inuitus: teneros ne tange poetas;
summoueo dotes impius ipse meas.
Callimachum fugito, non est inimicus amori;
et cum Callimacho tu quoque, Coe, noces.
me certe Sappho meliorem fecit amicae,
nec rigidos mores Teia Musa dedit.
carmina quis potuit tuto legisse Tibulli
uel tua, cuius opus Cynthia sola fuit?
quis poterit lecto durus discedere Gallo?
et mea nescioquid carmina tale sonant.

760

765

The erotic character of these poets works and hence their potential danger to the
patient of the Remedia is immediately spelled out in the adjective teneros (757: cf.

10 Note that such agricultural metaphors repeatedly carry sexual connotations in this didactic
poem: see generally E.W. Leach, Georgic Imagery in the Ars Amatoria, TAPA 95 (1964),
14254; Sharrock (n. 4), 2645; and cf. Robert Graves, Ovid in defeat, quoted in G. Liveley,
Ovid: Love Songs (London, 2005), 41 and ead., Ovid in Defeat? On the reception of Ovids Ars
Amatoria and Remedia Amoris, in Gibson et al. (n. 3), 31837 (Ovid instructs you how | Neighbours lands to plough, 1314).
11 For pectus = breast, see further OLD s.v. 1c. D.C. Feeney, To catch and to keep, The Times
Literary Supplement, 4 May 2007, 89 at 8 sees a comparable pun in Ars 1.7556, sunt diuersa
puellis | pectora.
12 Sharrock (n. 5), 49; contra, Janka (n. 5), 1245.
13 For a tabulated survey of correspondences, see A.A.R. Henderson, P. Ovidi Nasonis
Remedia Amoris (Edinburgh, 1979), xvi; cf. also Gibson (n. 6, [2007]), 1412.
14 So Coe, Rem. 760 ~ Coi poetae, Ars 3.329; Teia Musa, Rem. 762 / Ars 3.330; Tibullus and
Gallus by name, Rem. 763, 765 / Ars 3.334. There is nothing here, however, on Menander, Varro
or Virgil: contrast Ars 3.332, 3358.

SHORTER NOTES

283

teneri Properti, Ars 3.333).15 Then comes the Hellenistic pairing of Callimachus
and Philitas, the former ruled out as no foe to love (non inimicus amori, 759), and
the latter bracketed with him as one who does harm.16 Here the erotic implications
of this kind of reading matter are not probed any further, but in the next couplet
Ovids wry humour begins to show through: on 761 me meliorem fecit amicae,
Henderson, translating better company for my girl-friend, comments that the
adjective is inoffensive, but a nod is as good as a wink.17 Things get worse in the
following line, where the speaker declares that the erotic lyrics of Anacreon have not
led him into rigidos mores (762). Perhaps not, if we understand rigidus in its primary
sense here as strict or severe, and mores as morals. But in another sense the
development of rigidi mores, stiff habits,18 is precisely the kind of reaction we might
have expected the reading of erotic literature to produce and an equally good reason
for the lovesick to avoid it. Whatever physical or moral effects may (or may not) have
resulted from his reading of Anacreon, Ovids pun does little to dilute the sexual
content of his catalogue, and offers a further covert warning about the possible perils
of engaging with these suggestive texts.
A relatively innocuous line on Tibullus (763) leads into the next pentameter, which
is occupied by a periphrastic summary of Propertius elegiac poetry: cuius opus
Cynthia sola fuit (764). The inspiration for this description lies in Propertius own
declarations of exclusive devotion to Cynthia (see e.g. Prop. 1.12.20, 2.1.34,
2.13.356)19 but we should note that the older elegist nowhere refers to his beloved
as his opus. This refers primarily here to Propertius literary work, and at least his
first book of elegies was known by its opening word Cynthia (Prop. 1.1.1, 2.24.12;
Martial 14.189). In Roman elegy, however, and particularly in Ovid, who subjects the
erotic death fantasies of his elegiac predecessors to hilarious if tasteless literalization
by praying, at the mid-point of the Amores, that he may expire medium inter opus
(Amores 2.10.36), there is an all too easy slippage between opus, literary work, and
opus, sex.20 So when Ovid talks of the poet whose only job was Cynthia, he may be
giving a prcis of Propertius poetry, a summary of his sex life, or both. Once more the
characterization of this previous poets work is carefully chosen, and points yet again
to the impossibility of keeping sex off the brain when confronted with these
luminaries of erotic literature.
At line 765 Ovid moves on to C. Cornelius Gallus, the first major practitioner of
Latin love elegy. The proximity of the adjective durus to the name of this poet is
striking,21 given that duritia is the literary quality associated with Gallus elegies by
Quintilian in his catalogue of Roman elegists at Institutio oratoria 10.1.93:
15 Henderson (n. 13), 132; C. Lucke, P. Ovidius Naso Remedia Amoris: Kommentar zu Vers
397814 (Bonn, 1982), 330; P. Pinotti, Publio Ovidio Nasone: Remedia Amoris (Bologna, 1988),
320.
16 For Callimachus and Philitas as love poets, see also Prop. 2.34.312, 3.9.434; Ovid, Rem.
37982.
17 Henderson (n. 13), 132.
18 For rigidus used of erect genitalia, see OLD s.v. 3b; J.N. Adams, The Latin Sexual Vocabulary (London, 1982), 46, 103. Further Ovidian examples in J. Ingleheart, Burning Manuscripts:
the literary apologia in Ovids Tristia 2 and Vladimir Nabokovs On a Book Entitled Lolita ,
Classical and Modern Literature 26 (2006), 79109 at 858 with n. 29.
19 For the line-ending cf. Prop. 2.29B.24, in lecto Cynthia sola fuit: might recollection of lecto
in this Propertian context precondition the readers initial interpretation of the meaning of lecto
in Rem. 765 (see below)?
20 See e.g. D.F. Kennedy, The Arts of Love: Five Studies in the Discourse of Roman Love Elegy
(Cambridge, 1993), 5960; Ingleheart (n. 18), 101 with n. 69.

284

SHORTER NOTES

elegia quoque Graecos prouocamus, cuius mihi tersus atque elegans maxime uidetur auctor
Tibullus. sunt qui Propertium malint. Ouidius utroque lasciuior, sicut durior Gallus.

Could we see the appearance of durus in a Gallan context here in Ovid as evidence
that this critical judgement of the earlier poets work was already current before the
composition of Quintilians treatise? If so, the association of Gallus with duritia here
could be understood in one of two ways: either Ovid is defending his predecessor
from the charge later levelled against his poetry by Quintilian, claiming that it is
inconceivable that Gallus uvre could contaminate a reader with its own alleged
duritia, or he is slyly pointing to the humorous paradox of not coming away durus
oneself after reading a work notorious for its stylistic duritia. Might Quintilian even
have been influenced in his verdict by this line of Ovid? It is likely, moreover, that
Gallus himself used the word durus in a highly memorable and possibly programmatic
context, given the repeated occurrence of this adjective in what appear to be
imitations of Gallus own poetry.22 There may then be some playful engagement here
on Ovids part with terminology used by Gallus in his elegies: who could come away
hard-hearted (durus) after reading poetry traditionally characterized as tener, the
opposite of durus in its literary sense a distinction perhaps drawn by Gallus
himself ?23
However this may be, it should be noted that it is only with Gallo, the final word of
the hexameter, that the meaning of the line is actually fixed as wholl be able to go
away hard after reading Gallus? Before the clarification offered by this closing foot,
there is no indication (unless perhaps durus offers a hint along the lines suggested in
the preceding paragraph) that the line is about Gallus at all, and without the name at
the end quis poterit lecto durus discedere could mean something entirely different. For
lecto is revealed as a perfect passive participle, having been read, only when we reach
its complementary noun Gallo, whereas before this resolution the reader could be
forgiven for taking it as a substantive in its own right: wholl be able to leave the bed
(lectus) hard ?24 With his customary verbal dexterity, Ovid slips into his characterization of the emotional effects of reading Gallus an implicit pointer to the more
physical consequences of being propelled into sexual behaviour by this kind of poetry.
If you act on what you read in elegy, insinuates the praeceptor, theres no way youll
still be hard by the time you come to leave the bed: durus, like rigidus, can be used to
denote an erection.25 It is an entertaining deflation, perhaps literally so, when we find
out upon reaching the final word in the line that this is not in fact the sentences
primary meaning. Even then, however, there is room for doubt, given that gallus,
indistinguishable in manuscript from the name Gallus, can be used in Latin to denote
21

Some recentiores have tutus see Kenneys apparatus.

See Virgil, Ecl. 10.469, tu procul a patria (nec sit mihi credere tantum) | Alpinas, a! dura
niues et frigora Rheni | me sine sola uides. a, te ne frigora laedant! | a, tibi ne teneras glacies secet
aspera plantas!; Propertius 1.8.58, tune audire potes uesani murmura ponti | fortis et in dura naue
iacere potes? | tu pedibus teneris positas calcare pruinas, | tu potes insolitas, Cynthia, ferre niues?
(and, if I am right [PCPS 53 (2007), 16179 at 165], 3.7.478, nunc tulit et Paetus stridorem audire
procellae | et duro teneras laedere fune manus); F. Cairns, Sextus Propertius, the Augustan Elegist
(Cambridge, 2006), 8990, 111, 11415, 139, 191 (though here mollis is given as the opposite of
durus).
23 See passages in n. 22 above.
24 For discedere + simple ablative, see TLL 5.1.1280.3947.
25 J.T. Katz and K. Volk, Erotic hardening and softening in Vergils eighth Eclogue, CQ 56
(2006), 16974 at 1734 (Ecl. 8.801) with 173, n. 23 (Ecl. 4.30, after Nisbet); see also Ovid, Fasti
2.346, et tumidum cornu durius inguen erat.
22

SHORTER NOTES

285

a eunuch (specifically a devotee of Cybele: see e.g. Horace, Sat. 1.2.1202; Ovid, Fasti
4.3612; Martial 3.81, 11.74) or effeminate man.26 In this case the implication would
paradoxically (perhaps humorously) be one of amatory failure: who could come away
durus after reading someone whose very name might suggest sexual impotence?
Taken together, this extraordinary accumulation of innuendo over the course of
five lines is guaranteed, whatever else it does, to keep the readers mind on the subject
of sex. The dangers of reading works of erotic literature are themselves illustrated by
a reading of Ovids specification of the books to be avoided. No wonder he concludes
his catalogue with the observation et mea nescioquid carmina tale sonant (766)27 this
is after all precisely what he has been doing throughout the last few lines. And it is
what he does right the way through the Remedia, with the result that, as much recent
scholarship has observed,28 Ovids prescribed remedies are intrinsically programmed
to fail in their therapeutic capacity. In Rosatis words, Ovid well knows that amorous
speech (especially if made up from fragments from the great erotic poets) is a
symptom of latent illness and is itself a source of contagion: to continue to speak of
love even if only to repeat that it is over and that one is no longer in love means
continuing to dwell in the world of desire;29 for him, a line such as et mea nescio quid
carmina tale sonant () completing a list of erotic texts to be avoided sounds to the
reader of the Remedia like a confession (with an ironic wink) of the true nature not
only of the previous Ovidian elegy, but also of the Remedia itself (ibid.).
Both Rosati and Henderson, then, catch an Ovidian wink in the course of this
passage, at lines 761 and 766 respectively: in combination, those two instances might
be enough to prove the point. But if I am right about the verbal texture of the passage
as a whole, Ovid is not merely winking but positively fluttering his eyelashes.
University of Glasgow

L.B.T. HOUGHTON
l.houghton@classics.arts.gla.ac.uk
doi:10.1017/S00098388090000251

26 TLL 6.2.16867. For a possible pun along similar lines, see Prop. 1.10.12, primo cum testis
amori | adfueram, addressed to one Gallus (5): M. Pincus, Propertiuss Gallus and the erotics of
influence, Arethusa 37 (2004), 16596 at 1729.
27 CQs anonymous reader suggests there may be an echo here of Ovids provocatively reticent
cetera quis nescit? (Am. 1.5.25), ironically marking the avoidance of sexual explicitness in one of
the poets most sexual pieces (see esp. T. Schmitz, Cetera quis nescit. Verschwiegene Obsznitt
in der Liebesdichtung Ovids, Poetica 30 [1998], 31749; although Schmitz does not mention
Rem. 766, nescioquid tale would fit comfortably into his analysis).
28 See especially L. Fulkerson, Omnia vincit amor: why the Remedia fail, CQ 54 (2004),
21123; G. Rosati, The Art of Remedia Amoris: Unlearning to Love?, in Gibson, Green and
Sharrock (n. 3), 14365 at 1645; P.R. Hardie, Lethaeus Amor: The Art of Forgetting, in
Gibson, Green and Sharrock (n. 3), 16690 at 1667; Sharrock (n. 5), 623 (The rejection of love
is part of the discourse of love it is love, 62); ead., Ovid and the discourses of love: the
amatory works, in P.R. Hardie (ed.), The Cambridge Companion to Ovid (Cambridge, 2002),
15062 at 1601.
29 Rosati (n. 28), 164.

286

SHORTER NOTES

VIRGILS CUCUMBER AGAIN:

COLUMELLA 10.37892
Rebecca Armstrong has convincingly shown that the words tortusque per herbam |
cresceret in uentrem cucumis (Virg. G. 4.1212) lead the reader of Georgics 4 to expect
a snake, only to be surprised by the bathetic climax which reveals it to be a cucumber.1
This interpretation gains substantial support from an earlier reader, who signalled his
apprehension of Virgils suggestion by means of imitation. Columella, in his fifth
Georgic on horticulture, supplies the gap left for later authors by the invitation in the
time-pressed Virgils praeteritio.2 This is his description of the gourd and the
cucumber (10.37880, 38992):
tum modo dependens trichilis, modo more chelydri
sole sub aestiuo gelidas per graminis umbras
intortus cucumis praegnasque cucurbita serpit
liuidus at cucumis, grauida qui nascitur aluo
hirtus et ut coluber nodoso gramine tectus
uentre cubat flexo semper collectus in orbem,
noxius exacuit morbos aestatis iniquae.

As Armstrong has suggested with respect to the Virgilian passage, so commentators

have seen here in Columella an allusion to the serpentine name of the cucumis
anguineus.3 Yet, regardless or even because of this shared allusion, it is clear that
Columella is referring to the Virgilian passage and, by explicitly comparing the
cucumber to two different snakes, is signalling his apprehension of the latters strategy
of suggesting but suppressing mention of a snake.4 He marks the allusion to Virgils
cucumber by recalling almost every word in the Georgics passage, either by verbal
echo (tortus ~ intortus, uentrem ~ uentre) or by the use of synonyms and periphrases
(tortus ~ flexo,5 per herbam ~ per graminis umbras), while the notion of cresceret is less

R. Armstrong, Virgils cucumber: Georgics 4.1212, CQ 58 (2008), 3668.

Virg. G. 4.1478, explicitly invoked at Col. 10.pr.3, 15. Columella appointed himself Virgils
heir and stepped into a breach that did not really exist. (E. Gowers, Vegetable love: Virgil,
Columella, and garden poetry, Ramus 29 [2000], 12748, at 127). The most sympathetic and
stimulating discussions of Col. 10 are Gowers, J. Henderson, Columellas living hedge: the
Roman gardening book, JRS 92 (2002), 11033, V.E. Pagn, Rome and the Literature of Gardens
(London, 2006), 1936, and S. Diederich, Rmische Agrarhandbcher zwischen Fachwissenschaft,
Literatur und Ideologie (Berlin, 2007), 22758.
3 Armstrong (n. 1), 367. F. Boldrer (ed.), L. Iuni Moderati Columellae Rei Rusticae Liber
decimus (Carmen De Cultu Hortorum) (Pisa, 1996), 328, ad 10.378, who adds that the amphibious chelydrus is an effective parallel for the water-loving cucumber. The pun may also lie behind
Martials intriguing connection of reptile and vegetable in describing a farm so tiny that a
cucumber cannot lie straight nor a whole snake fit in it (11.18.1011).
4 Hendersons ludic translation of this passage suggests, with its suspenseful dots, exactly the
sort of bathetic surprise which Armstrong finds in Virgil: Here, hanging in an outhouse, or there,
like some watersnake, | out in the summer sun, all through the cool shade of the grass, | there
creeps cucumber, J. Henderson, The Roman Book of Gardening (London, 2004), 63. Of course
Columella, unlike Virgil, has already made explicit that whatever he is describing is like, and
hence is itself not a snake, but Henderson brings out the misdirection which Columella replicates
from and annotates in Georgics 4.
5 Columellas flexo also echoes flexi uimen acanthi in Virgils next line at G. 4.123. It is
notable that Columella swaps the epithets by referring to tortos acanthos earlier at 10.241, the
allusion marked by an Alexandrian footnote since the artichoke imitatur the acanthus. Gowers
2

SHORTER NOTES

287

directly evoked by the conventional image of the pregnant gourd and the boldly
paradoxical one of the cucumber being born from its own laden womb.6 The extreme
brevity of the similes (more chelydri, ut coluber) underlines the close parallelism
between snake and cucumber, since precisely the same language can be used of both
with no need for differentiation between tenor and vehicle. In fact, one might go
further and argue that the language is rather more appropriate to the snakes than to
the cucumber. This would constitute an inversion of what Oliver Lyne has termed
trespass, so that here words like serpit, uentre and cubat stray from the (implied)
simile into the narrative.7 This further reinforces Columellas implicit assertion that
Virgils cucumber is similarly described in terms appropriate to a snake, but with the
simile marker ut coluber suppressed.
Columella demonstrates his awareness of the snakiness of Virgils cucumber, not
only by allusion to the passage itself, but by combining that with allusion to the actual
snakes described in Georgics 3 and elsewhere. In this way, as Hardie in particular has
demonstrated with authors such as Ovid and Valerius Flaccus, combinatorial
imitation of two separate sections of Virgils text can illuminate the connection
between them.8 Thus the two snakes to which Columella compares the cucumber (and
in the first case also the gourd) are the chelydrus and the coluber, the same two snakes
named at G. 3.415 and 418. Schroeter even suggests ingeniously that the cadence more
chelydri might be shaped in imitation of the latter lines nidore chelydros.9 Even more
suggestively, 10.391 echoes, as Brakman notes, Virgils description of the monstrous
serpents which, in the laudes Italiae, are absent from Italy: nec rapit immensos orbis per
humum neque tanto | squameus in spiram tractu se colligit anguis (Virg. G. 2.1534).10
This reference to the (problematic) absence of snakes is particularly suggestive in a
context where Columella is drawing attention to the unexpected absence of a snake in
Virgils cucumber patch.11
(n. 2), notes how the use of imitatur (the sand imitates the rope) at 10.8 prefigures the imitative
mode of the whole poem. On divided allusion, see J. Wills, Divided allusion: Virgil and the
Coma Berenices, HSCPh 98 (1998), 277305.
6 The allusion of the whole passage to G. 4.1212 is noted by Boldrer (n. 3), 329, ad 10.379b.
The parallel of intortus cucumis to tortus cucumis is also noted by G. Schroeter, De Columella
Vergilii Imitatore (Jena, 1882), 37, E. de Saint-Denis, Rhabilitons Columelle pote, GIF 21
(1969), 12136, at 124, and A. Biotti (ed.), Virgilio, Georgiche Libro IV (Bologna, 1994), 116, ad
Virg. G. 4.1212. In giving the cucumber as well as the gourd a pregnant and hence distended
belly, Columella sides with Virgil against (or perhaps tries to reconcile him with) his correctors
at Prop. 4.2.43 (noted by Armstrong [n. 1], 368) and Mor. 75 (noted by CQs anonymous reader).
7 On trespass, see R.O.A.M. Lyne, Words and the Poet (Oxford, 1989), 734, and see index s.v.
trespass. Boldrer (n . 3), 329, ad Col. 10.380 refers to serpit as a verbo espressivo che anima
lortaggio e prosegue la similitudine con un serpente.
8 P.R. Hardie, Flavian epicists on Virgils epic technique, Ramus 18 (1990), 320; id., The Epic
Successors of Virgil (Cambridge, 1993), 14, 43, 47, 789. Particularly relevant is his observation
that Ovids allusion, in describing the serpent killed by Cadmus in Met. 3, to both the serpents
which kill Laocon in Virg. A. 2 and to Cacus in A. 8, reveals an alertness to significant structural correspondences in [the Aeneid], (id. Ovids Theban history: the first anti-Aeneid?, CQ
40 [1990], 22435, at 227). On the contaminatio in Col. 10 of allusions to different parts of the
Georgics, as well as the other Virgilian poems, see Boldrer (n. 3), 21.
9 Schroeter (n. 6), 37.
10 C. Brakman, Ad Columellae librum decimum, Mnemosyne 60 (1933), 10712, at 111.
Boldrer (n. 3), 332, ad 10.391 does not mention this echo but suggestively (for both Virgil and
Columella) notes the verbal parallels with descriptions of monsters at Cic. Arat. fr. 8.3 T and
Prop. 4.6.35.
11 CQs anonymous reader suggests a further instance of combinatorial imitation. The passage
immediately preceding (10.35768), describes the ritual expulsion of caterpillars (dira

288

SHORTER NOTES

Columella is not, however, merely displaying his doctrina by detecting the subtle
allusion to the snake in the Georgics and creatively annotating it through explicit
similes. His imitation also constitutes an interpretation.12 Armstrong concludes her
discussion with some suggestions as to how the substitution of the cucumber for the
anticipated snake might form a small-scale comment on some of the larger themes of
the poem.13 In her reading the very suppression of the expected snake is not a
suggestion that the apparently harmless cucumber has sinister qualities, but rather a
demonstration that not all surprises lurking in the grass of the Georgics are nasty;
some can be welcome: the plump cucumber thus adds to the sense that a garden has
the potential to surprise with its bounty.14 Similarly, Henderson, without noting
Virgils serpentine subtext, contrasts the two depictions: Columellas cucumber, long
and round, serves up a sinister deformity for the fall, where Virgils garden saw only a
joy he could have grown into a greenhouse of delight, no worries.15
As always with such creative allusion, two broad possibilities are available to
the reader. One might argue that Columella polemically puts the snake back into the
cucumber; in doing so, he dissents from Virgils suppression of the serpent and his
related implication that the cucumber is a harmless, even desirable vegetable, thus
further reinforcing Columellas own negative, or at least ambivalent, depiction of this
vegetable.16 Alternatively, Columellas allusion might support a reading whereby the
implicit snakiness of Virgils cucumber was in itself a subtle indication of its sinister
potential within the world of the Georgics, an indication which the imitator and
commentator draws attention to, emphasizes and validates. This latter possibility
might be reinforced by the combinatorial imitation, noted above, of Virgils snakes
and Virgils cucumber. Indeed, the reference to the cucumbers harmfully exaceranimalia! 351 nesso epicheggiante ed iperbolico, Boldrer [n. 3], 316) from the garden by a
bare-footed (nuda plantas), menstruating woman, and compares her to Medea sending the
Colchian serpent to sleep. As well as being another example of Columellas explicit paralleling of
a relatively minor pest with a monstrous serpent, the reader suggests that this scene might recall
the real serpent of Georgics 4, the immanis hydrus lying unseen in the long grass before Eurydices
(presumably, but not explicitly, bare) feet (4.4579). This is attractive, though the emphatically
ritual context of this pot pourri of mumbo jumbo (Henderson [n. 2], 130), in which even the
bare feet have magical associations (Boldrer [n. 3], 319), and the close verbal echo of Senecas and
Lucans Medeas (magicis cantibus, Col. 10.367, Med. 684, Luc. 4.553) make the evocation of
Eurydice a secondary one.
12 On Columella as a tool for interpreting Virgil, see esp. E. de Saint-Denis, Columelle, miroir
de Virgile, in H. Bardon (ed.), Vergiliana: recherches sur Virgile (Leiden, 1971), 32843. Where
Saint-Denis studies ce que luvre de Columelle nous rvle sur celle de Virgile, telle quelle a t
lue et comprise au premier sicle de notre re (329) and argues that parce que Columelle est plein
de Virgile et quil a vcu en communion intime avec lui, nous pouvons linterroger pour essayer de
rsoudre quelques nigmes irritantes et de mettre un point final des controverses interminables
(337), one might prefer to see Columella offering tendentious interpretations of, rather than
absolute solutions to, the enigmas of Virgils texts. Cf. Gowers (n. 2), 127: it tells us how Virgil
was read in antiquity, and Diederich (n. 2), 231: Man hat festgestellt, da Columellas
Vergilimitation, komplex und subtil wie sie ist, viel Kreativitt verrt. On references to Virgil
outside Book 10, see A. Cossarini, Aspetti di Virgilio in Columella, Prometheus 3 (1977),
22540; A. Doody, Virgil the farmer? Critiques of the Georgics in Columella and Pliny, CPh 102
(2007), 18097.
13 Armstrong (n. 1), 367.
14 Ibid. 368.
15 Henderson (n. 2), 130, n. 52.
16 On the cucumbers (and the gourds) ambivalence, see Henderson (n. 2), 130: this pair of
fat-bellies, everything from fatal through life-saving, as versatile as Columellas verse, as necessary
as his expertise.

289

SHORTER NOTES

bating the diseases of the harsh summer (10.392) might take the combinatorial
imitation further in pointing the imagistic parallel between the snakes and the plague
which are the two principal destructive forces of the second half of Georgics 3.17
Certainly the plague and the cucumber both slither (serpant, G. 3.475), and exercise
their destructive power in summer (aestiua, 3.472).
Ultimately, however, Columellas world view, like his cucumber, is a complex
mixture of (in crude terms) positive and negative.18 The summer brings horticultural
bounty as well as the blasting Dog Star (10.4008); the serpentine cucumis liuidus may
aggravate disease, but the cucurbita can be used as a wine-container or a buoyancy aid
for children learning to swim (3878), while the cucumis candidus will not only soothe,
but actively bring aid to the ill (3949).19 Columella certainly identifies the snakiness
of Virgils cucumber and it is more than arguable that his annotative imitation also
interprets it, but this is not to say that he asserts a simplistic or reductive interpretation which would diminish the glorious complexity and ambiguity of Virgils
Georgics, Columellas De Cultu Hortorum and both of their cucumbers.
Balliol College, Oxford

ROBERT COWAN
bob.cowan@balliol.ox.ac.uk
doi:10.1017/S00098388090000263

17 On parallels between Virgils snakes and plague, see, inter alios, D.A. Ross, Virgils Elements
(Princeton, 1986), 17783; R.F. Thomas, Virgils Georgics vol. 2 (Cambridge, 1988), 119, ad
3.41439.
18 Columella erkennt die Ambivalenz der Natur also durchaus an und huldigt keinem so
einseitigen realittsfernen Optimismus, wie ihm die Forschung vielfach vorgeworfen hat,
Diederich (n. 2), 245. On the cucumber, see Henderson quoted in n. 16 above. With different
emphasis, P. Toohey, Epic Lessons (London, 1996), 1769, notes a tension in the poem between
erotic (and fertile) sensuality and chaste purity. On the associations of cucurbits with the imagery
of fertility in literature more broadly, see R. Norrman and J. Haarberg, Nature and Language: a
Semiotic Study of Cucurbits in Literature (London, 1980), 1379, esp. 213 on classical texts.
19 I use Columellas own terms to distinguish between the cucurbits, but they are notoriously
difficult to identify. For a scientific attempt to do so, see J. Janick, H.S. Paris and D.C. Parrish,
The cucurbits of Mediterranean antiquity: identification of taxa from ancient images and
descriptions, Annals of Botany 100 (2007), 144157, at 14445. Rebecca Armstrong suggests per
litteras that Columellas observation on the diversity of the cucurbits, una neque est illis facies
(10.381), might further constitute a correction of Virgils failure to catalogue the different kinds. I
am grateful to her and to CQs anonymous reader for their helpful suggestions.

THEMIS AT ELEUSIS: CLEMENT OF

ALEXANDRIA, PROTREPTICUS 2.22.5
The present note argues against a tenacious emendation. The textual problem with
which it is concerned is located in a particularly sensitive passage, the long attack
against the Mysteries of the Hellenes in Clement of Alexandrias Protrepticus 2.122,
one of our most precious sources of information on ancient mystery cults. The
manuscripts of the Protrepticus are unanimous in reading
at 2.22.5:1
1 All extant manuscripts of the Protrepticus derive from the Parisinus Graecus 451 (P), which
dates from the tenth century. See O. Sthlin, Clemens Alexandrinus. Bd. 1: Protrepticus und
Paedagogus (Leipzig, 1905), xvixxiii.

290

SHORTER NOTES

< >
The passage is cited by Eusebius with the exact same words.2 The emendation proposed by Wilamowitz in 1880, however, correcting the
of the passage
into
, has met with almost unanimous approval.3 It has been used by
2 Eusebius transcribes this section of the Protrepticus (2.11.123.9) in his Praeparatio
Evangelica 2.3.40; K. Mras, Eusebius Werke. Praeparatio Evangelica, vol. 1 (Berlin, 19542), 86.
Although Wilamowitz writes corruptela autorem etiam Eusebii exemplum obsedit, the manuscript
reading
was retained by Mras in his 1954 edition of the Praeparatio Evangelica. All MSS of
Eusebius PE, together with P, are unanimous in reading
, and even Wilamowitz concedes that
this is what Eusebius also read (from a corrupted copy, in his opinion). Eus. BONV, together with
Clem. P, all agree in reading
after
. Eus. H provides the only discordant reading:
(M. Marcovich, Clementis Alexandrini Protrepticus [Leiden, 1995], 32). Arnobius makes
a free translation of Clements passage in Adv. Nat. 5.19.26, but without any echo of the lines on
Themis.
3 U. von Wilamowitz-Mllendorff, Commentariolum Grammaticum, vol. 2 (Gttingen, 1880),
11 = U. von Wilamowitz-Mllendorff, Kleine Schriften IV: Lesefrchte und Verwandtes, ed.
K. Latte (Berlin, 1962), 6089. Until recently, Clements discussion of the mysteries in the
Protrepticus was widely seen as an aggregate of information from various telestic cults, arranged
without much order, and the case for seeking an appropriate divinity from any other mystery cult
to make sense of the passage seemed perfectly legitimate. In 1880 there was no indication that
Themis might have had an established presence in the Eleusinian cult, or that she played a part in
telestic rites more generally. Lobeck, for instance, already suggested the correction
(C.A. Lobeck, Aglaophamus, sive de theologiae mysticae Graecorum censis libri tres [Knigsberg,
1829], 703). Mras conjectures
on the basis of the
of Eus. H. Themis did not
seem to fit a context of teletai, and G Themis was a brilliant choice for an emendation. The idea
of Themis as the expression of a primordial earth-goddess has had long currency; see e.g.
L.R. Farnell, The Cults of the Greek States, vol. 3 (Chicago, 1971), 3, 13; K. Latte, Themis, RE2
5.2 (1934), 16267; K. Reinhardt, Personifikation und Allegorie, Vermchtnis der Antike:
gesammelte Essays zur Philosophie und Geschichtsschreibung, ed. K. Becher (Gttingen, 19662),
26; K. Clinton, IG i2. 5, the Eleusinia, and the Eleusinians, AJP 100 (1979), 7; R.F. Healey,
Eleusinian Sacrifices in the Athenian Law Code (New York, 1990), 759; 21924. An equivalence
between Ge and Themis is made by Aeschylus, PV 2102, for instance (cf. A.J. Podlecki,
Aeschylus. Prometheus Bound [Oxford, 2005] ad loc.); she is associated with Earth at Delphi (E.
Stafford, Worshipping Virtues: Personification and the Divine in Ancient Greece [London, 2000],
526); the inscription of a seat in the theatre of Dionysus at Athens (IG 22 5130) reads
(cf. also IG 22 5098: two hersephoroi of Chloe Themis); Paus. 5.14.10 sets the altar of
Themis in the mouth of the oracle of Ge, and locates a sanctuary of Themis next to a sanctuary
of Ge on the south slope of the Acropolis (1.22.13; cf. 8.25.7). The understanding of Themis as
the personification of an original deity of fertility and social order belongs to a distinct moment
in the history of religions, of course. Stafford, 4573, the most recent scholarly discussion of
Themis as a figure of cult, rightly warns against continuing to privilege the early and primitive
association of Themis with earth and fertility. As all divinities of polytheistic Greek religion,
Themis is linked to many other gods through rich ties of equivalence, complementarities and
proximity, and there is no valid reason to privilege the tenuous connection of Themis with Ge in
archaic and classical (mostly) literary sources to the detriment of all other associations (see e.g.
Delphi inv. 4286; H.A. Shapiro, Personifications in Greek Art: The Representation of Abstract
Concepts 600400 BC [Zurich, 1993], no. 144). Not only is the cult association between Ge and
Themis not original, but it is hardly attested. No actual cult association between Themis and Ge
is attested before the mid-third century B.C.E., as Stafford observes, and even there the few
documents adduced are far from decisive (67: nowhere have we seen evidence to support the
theory that Themis originated as an epithet of earth). Price (T.H. Price, Kourotrophos. Cults and
Representations of the Greek Nursing Deities [Leiden, 1978], 10132), for instance, has argued
that the inscription which was thought to mention a priestess of Ge Themis (IG 22 5130) might
actually be referring to a priestess of Ge and Themis instead, or even to two distinct priestesses,
one of Ge and one of Themis (see also Stafford, 634). This late imperial text is in fact our only

SHORTER NOTES

291

some to dissociate the passage entirely from Clements description of Eleusis.4 Others
have seen it as pointing towards an equivalence between Demeter and Themis, as yet
another generic type of earth-mother goddess, whether located in Eleusis or not.5
Although this emendation has been convincingly refuted by Pierre Boyanc in 1936, it
continues to be generally accepted in editions of the text and discussions of the
passage, when it is recognized as an emendation.6 As the strong case for the
of the manuscripts has all but been ignored in the past 70 years, and as our
understanding of Clements sources on mystery cults has significantly increased
during that time, we believe that the question needs to be revisited. This is a textbook
example of a hermeneutic house of cards: layers of interpretation built on the single
(virtual) stroke of one letter.7
Our passage culminates with an assault on the Mysteries par excellence, the
Eleusinian rites (2.202).8 After revealing and exposing a number of other
mysteries (Aphrodite, Deo, Dionysus, Corybantes), Clement moves on to attack the

independent attestation for the putative figure of Ge-Themis (C. Sourvinou-Inwood, Myth as
history: the previous owners of the Delphic Oracle, in J.N. Bremmer [ed.], Interpretations of
Greek Mythology [Oxford, 1987], 240, n. 62; Stafford, 68).
4 See e.g. A. Krte, Zu den eleusinischen Mysterien, ArchRW 18 (1915), 11525; L. Deubner,
Attische Feste (Berlin, 1932), 82; U. Pestalozza, Religione mediterranea (Milan, 1951), 22933; G.
Casadio, Vie gnostiche allimmortalit (Brescia, 1997), 1966. Most scholars who have written on
the question do not mention that
is an emendation.
5 Krte (n. 4), for instance, argues that the kteis gunaikeios was shown to the initiand as a
symbol of rebirth. He writes that Clement called Demeter GeThemis as er nur der Abwechslung halber die eleusinische Erdmutter mit einem anderen, wesengleichen Namen benennt.
Pestalozza (n. 4), 231 makes GeThemis an equivalent of DemeterKore. Cf. also V. Ehrenberg,
Die Rechtsidee im Frhen Griechentum (Darmstadt, 1966), 33; Farnell (n. 3), 1314.
6 P. Boyanc, Le culte des Muses chez les philosophes grecs (Paris, 1936), 36, 53; P. Boyanc,
Sur les mystres dleusis, REG 75 (1962), 46082. The editions of O. Sthlin (n. 1), G.W.
Butterworth, The Exhortation to the Greeks, The Rich Mans Salvation, and the Fragment of an
Address Entitled To the Newly Baptized by Clement of Alexandria (Harvard, 1919), Q. Cataudella,
Clemente Alessandrino. Protreptico ai Greci (Turin, 19402), and M. Marcovich (n. 2), the
most recent one, all follow Wilamowitzs opinion. C. Mondsert, Clment dAlexandrie. Le
Protreptique (Paris, 1949), is alone in accepting the
of the manuscripts. Roberts also
argues against the emendation of Wilamowitz, on the basis that Clement almost never
hyphenates the name of a god or goddess (L. Roberts, The unutterable symbols of ( ),
HThR 68 [1975], 79). His discussion is an example of the force of scholarly tradition, as it
continues to presuppose the association of Themis with Ge in the passage. See e.g. p. 79: the
Alexandrian procession may have included three different sets of mysteries. This does not
discount the possibility that Clement may have combined the two goddesses in his thought. For
scholars now recognize that the ancient goddess
remained important through Hellenistic
times. The persistent presence of GeThemis in discussions of the passage, even when the
emendation of Wilamowitz is not accepted, is striking. Like Roberts, for instance, Stafford (n. 3),
4573 accepts the manuscript reading
, but explicitly reads the passage in reference
to GeThemis. Although A. Bernab, Orphicorum et Orphicis similium testimonia et fragmenta.
Poetae epici graeci. Pars II, 3 vols (Leipzig, 20047), rightly edits
in OF 532 II, he
locates the fragment in the section devoted to Phlya et Lycomidae; he is following Casadio
(n. 4), who actually bases this link with Phlya on the reading
.
7 The difference between
and is of course minimal in uncial scripts, and a mistake readily
made.
8 On Clements view of Eleusis, see for instance F. Graf, Eleusis und die orphische Dichtung
Athens in vorhellenistischer Zeit (Berlin, 1974), 1949; W. Burkert, Homo Necans: The Anthropology of Ancient Greek Sacrificial Ritual and Myth, trans. P. Bing (Berkeley, 1983), 251, 26971.
Wilamowitz had a generally low opinion of Clements value as witness for the Eleusinian cult:
U. von Wilamowitz-Mllendorff, Der Glaube der Hellenen, vol. 2 (Berlin, 1932), 374.

292

SHORTER NOTES

mysteries of Pherephatta.9 Throughout the passage, Clement systematically relates

the myth of each mystery cult which is being exposed, and proceeds to reveal its
synthmata, hagia, and/or symbola. The section on the Cabiri, for instance, ends with
a description of the shameful symbolon of the cults mystic kist: the castrated aidoion
of Dionysus (2.19.14).10 In disclosing the myths and rites of Eleusis as so many
empty errors and shocking obscenities, the attack against the mysteries of
Pherephatta follows the same pattern deployed against other teletai in the passage. It
also follows the same order of material presentation. It tells the myth of Demeters
search for her daughter in the version of a hymn attributed to Orpheus (2.20), quotes
some verses from Orpheus which illustrate the shamelessness of the cult (2.21.1),
discloses the
(2.21.2), and reveals the
of the
mystic kistai (2.22.4). Immediately after this, Clement proceeds to unveil the
of the cult (2.22.5).
Wilamowitzs emendation was based on internal and external considerations. The
passage of Clement was thought to have little order in its arrangement of the material,
something which justified looking beyond Attic Eleusis for fitting parallels;11
moreover, no independent evidence for Themis and the mysteries was thought to exist,
something which justified changing the received text.12 Both considerations have now
been disproved. First, the sentence which mentions the unutterable symbols of
Themis clearly belongs to Clements discussion of Eleusis. Recent research shows that
the arrangement of the material we find in this section follows the pattern used for the
other cults, as we have seen.13 The mention of the symbola at 2.22.5 belongs squarely
in the section of the passage concerned with Eleusis.14 The reference to the Hierophant, the Dadouch, and Iakchos right after the aporrhta symbola of Themis are
9 The systematic alphabetical order adopted in the passage makes it clear that Clement is
following a written source, a Hellenistic treatise on the mysteries which gives mythical and ritual
accounts of diverse teletai using as main source one (or several) Orphic poem(s); see
C. Riedweg, Mysterienterminologie bei Plato, Philo und Klemens von Alexandrien (Berlin, 1987);
N. Robertson, New Light on Demeters Mysteries: The Festival Proerosia, GRBS 37 (1996),
36575; M. Herrero, Las fuentes de Clem. Alex. Protr. 2.1222: un tratado sobre los misterios y
una teogona rfica, Emerita 75 (2007), 3750. This contradicts the obviously interested information of Eusebius (PE 2.2.64) that Clement had direct knowledge of the mysteries because he
had been initiated, something which is still widely repeated. While with Pherephatta Clement
designates Kore, the poetic epiklsis Deo is used to refer to both Rhea and Demeter. The alphabetical treatise used by Clement probably relies on Orphic poetic traditions in which these
equations of feminine goddesses were typical (e.g. P. Derveni, col. XXII.7 [OF 398 Bernab]).
10 On the meaning of kist in the passage, see Roberts (n. 6), 779.
11 Roberts (n. 6) 73, for instance, still sees the passage concerning the symbola of Themis as an
afterthought of Clements general discussion of the mysteries.
12 Neither P.F. Foucart, Les mystres dleusis (Paris, 1914), G. Mylonas, Eleusis and the
Eleusinian Mysteries (Princeton, 1961), K. Kernyi, Eleusis: Archetypal Image of Mother and
Daughter, trans. R. Manheim (New York, 1967), K. Clinton, The Sacred Officials of the
Eleusinian Mysteries (Philadelphia, 1974), nor K. Clinton, Myth and Cult: The Iconography of
the Eleusinian Mysteries (Stockholm, 1992), for instance, discuss the presence of Themis at
Eleusis. Healey (n. 3), 7583, explicitly describes Themis as a divinity who does not belong to the
Eleusinian pantheon.
13 See n. 9.
14 The fact that Clement breaks his unveiling of Eleusis with two rhetorical interruptions
(2.22.12 and 2.22.3, based on Heraclitus and Philo, respectively), and that he has also anticipated some attacks on Eleusis in 2.12.2 and 2.17.1, does not affect the validity of this statement.
The important point is that we expect to find symbola at the end of the section. A new lemma
with Themis would also be uncharacteristic in both having no myth and breaking the alphabetical order.

SHORTER NOTES

293

mentioned should be sufficient to confirm the matter.15 Second, the external considerations which were valid in 1880 have changed since then. It is no longer true to say,
with Wilamowitz, that ignoramus Themidis mysticum cultum.16 The publication of
column III from the Nichomachean sacrificial calendar in 1935 confirmed the
presence of Themis in the rituals of Eleusis once and for all.17 Whereas Themis was
nowhere associated with the Eleusinian cult of Demeter and Kor before this text was
found, the sacrificial calendar of the Agora now gives her pride of place in the rituals
of the mysteries at the end of the fifth century B.C.E. The nineteen sacrifices
mentioned in the upper part of column III are to be performed by the Eumolpids in
the Eleusinia festival.18 The very first sacrifice mentioned is a ewe of twelve drachmas
for Themis.19
The presence of Themis in the cult of Eleusis is independently and prominently
attested. There is no valid reason to continue defending the emendation of
Wilamowitz against the unanimous reading of the manuscripts and of Eusebius. The
lemma of the Protrepticus (and of its source) on the mysteries of Eleusis mentions
and describes the
of Themis.20 Whatever value one gives to this
testimony, it is a significant element of the great Eleusinian puzzle. It can no longer be
replaced by a virtual text.
McGill University
Universit di Bologna

RENAUD GAGN
renaud.gagne@mcgill.ca
MIGUEL HERRERO
miguelhdj@yahoo.es
doi:10.1017/S00098388090000275

15 The rhetorical invocation which follows our passage has been thoroughly analysed in
C. Riedweg, Die Mysterien von Eleusis in rhetorisch geprgten Texten des 2/3. Jahrhunderts
nach Christus, Ill. Class. Stud. 13.1 (1988), 12733.
16 Wilamowitz (n. 3), 11 = 6089.
17 F. Sokolowski, Lois sacres des cits grecques. Supplment (Paris, 1962), n. 10, l. 60, first
published in J.H. Oliver, S. Dow, Greek Inscriptions, Hesperia 4 (1935), 219. See now
S. Lambert, The sacrificial calendar of Athens, ABSA 97 (2002), 35399. Boyanc (n. 6, 1936),
26 and 53 and (n. 6, 1962), 4802, was the first to see the importance of this inscription for the
passage of the Protrepticus. He, however, improbably connects the Eleusinian Themis with the
of Proclus, In rempubl. 1.125.20 Kr. Healey (n. 3), 7583, discusses the presence of
Themis in the inscription at length. His contention that the sacrifice of Themis mentioned in this
public inscription belongs to the gentilitial cult of the Eumolpidae, based on the understanding
of la thmis ancestrale as the symbol of clan justice proposed by G. Glotz, La solidarit de la
famille dans le droit criminel en Grce (Paris, 1904), and his many followers long ago, no longer
carries conviction.
18 For the identification of this list as referring to the Eumolpid sacrifices of the Eleusinia
festival, and not the Mysteries themselves, see Lambert (n. 17), contra Oliver (n. 17), 269 and
Sokolowski (n. 17) ad loc.
19 Lambert (n. 17), l. 60:
. The price is wrongly indicated in Sokolowski. For
discussion of sacrificial prices in the inscription, see Lambert (n. 17), 3967.
20 Marjoram, a lamp, a sword and a womans comb are mentioned, probably objects to be
placed in a sacred basket, as Stafford (n. 3), 625 suggests, in reference to the hersephoroi of IG 22
5098. This short discussion on the emendation of Wilamowitz is not the place to speculate on the
significance of this group of objects.

THE CLASSICAL ASSOCIATION

The Classical Association has a worldwide membership and is open to all who
value the study of the languages, literature, and civilizations of ancient Greece
and Rome. It creates opportunities for friendly exchange and co-operation
among classicists, encourages scholarship through its journals and other
publications, and supports classics in schools and universities. Every year it
holds an annual conference, and it sponsors branches all over the country
which put on programmes of lectures and other activities.
The Classical Association has about 3,500 members. Members receive The
Presidential Address once a year and a newsletter, CA News, twice a year. They
may also subscribe at substantially reduced cost to the Classical Association
journals Classical Quarterly, Classical Review, and Greece & Rome.
For more information and a membership application form please contact
the Secretary, The Classical Association, Senate House, Malet Street,
London WC1E 7HU, UK (Tel: +44 (0)20 7862 8706, email:
office@classicalassociation.org). The Secretary can also give information
about the reduced journal subscription rate, and about the Classical
Associations other occasional publications.
THE CLASSICAL QUARTERLY
The Classical Quarterly has a reputation for publishing the highest quality
classical scholarship for nearly 100 years. It publishes research papers and
short notes in the fields of language, literature, history, and philosophy. Two
substantial issues (around 300 pages each) of The Classical Quarterly appear
each year, in May and December. Given the quality and depth of the articles
published in The Classical Quarterly, any serious classical library needs to have
a copy on its shelves.

Typeset by Forewords, Oxford

Printed by Bell & Bain Ltd, Glasgow

NEW SERIES

THE CLASSICAL QUARTERLY

C
THE
CLASSICAL
QUARTERLY

VOLUME 59 NUMBER 1 2009

Cambridge Journals Online

For further information about this journal please
go to the journal web site at:

journals.cambridge.org/caq

PUBLISHED FOR THE CLASSICAL ASSOCIATION

BY CAMBRIDGE UNIVERSITY PRESS